The Revised Body Awareness Rating Questionnaire: Development Into a Unidimensional Scale Using Rasch Analysis

The Revised Body Awareness Rating Questionnaire: Development Into a Unidimensional Scale Using... Abstract Background The Body Awareness Rating Questionnaire (BARQ) is a self-report questionnaire aimed at capturing how people with long-lasting musculoskeletal pain reflect on their own body awareness. Methods based on classical test theory were applied to the development of the instrument and resulted in 4 subscales. However, the scales were not correlated, and construct validity might be questioned. Objective The primary purpose of this study was to explore the possibility of developing a unidimensional scale from items initially collected for the BARQ using Rasch analysis. A secondary purpose was to investigate the test-retest reliability of a revised version of the BARQ. Design This was a methodological study. Methods Rasch and reliability analyses were performed for 3 samples of participants with long-lasting musculoskeletal pain. The first Rasch analysis was carried out on 66 items generated for the original BARQ and scored by 300 participants. The items supported by the first analysis were scored by a new group of 127 participants and analyzed in a second Rasch analysis. For the test-retest reliability analysis, 48 participants scored the revised BARQ items twice within 1 week. Results The 2-step Rasch analysis resulted in a unidimensional 12-item revised version of the BARQ with a 4-point response scale (scores from 0 to 36). It showed a good fit to the Rasch model, with acceptable internal consistency, satisfactory fit residuals, and no disordered thresholds. Test-retest reliability was high, with an intraclass correlation coefficient of .83 (95% CI = .71–.89) and a smallest detectable change of 6.3 points. Limitations The small sample size in the second Rasch analysis was a study limitation. Conclusions The revised BARQ is a unidimensional and feasible measurement of body awareness, recommended for use in the context of body-mind physical therapy approaches for musculoskeletal conditions. Body awareness is an overall concept of body consciousness, body management and body experience.1 The phenomenon, concerns the experience of coherence between body and emotions, the ability to sense and being confidential with habitual reactions of the body as well as attitudes toward it. Body awareness seems to be a prerequisite for self-experience.2 The phenomenon as addressed in physical therapy deals with the patient's ability to sense muscle tension and body movements, reflecting emotions and attitudes toward own body. Such ability seems to be a prerequisite for readjusting posture and muscle tension in treatment.2–4 The phenomenon of body awareness is accordingly a core aspect addressed in body–mind treatment approaches like Norwegian Psychomotor Physiotherapy (NPMP)2–4 and Basic Body Awareness Therapy (BBAT)5,6 as well as in other treatment modalities used by physical therapists working in the field of musculoskeletal and mental problems. To properly assess and evaluate body awareness in patients with long-lasting musculoskeletal pain, physical therapists need a reliable and valid measurement tool that is feasible for clinical use. The questionnaires retrieved from search in the literature assessed only related constructs like body image and attitude toward own body,7–10 and no appropriate tool was found available to assess body awareness in the context of physical therapy. Therefore a new self-reported questionnaire was developed, the Body Awareness Rating Questionnaire (BARQ).11,12 The BARQ was aimed to capture body awareness among patients with long-lasting musculoskeletal pain. Items considered to reflect body awareness were collected among NPMP therapists and from descriptions among patients receiving or being on a waiting list for NPMP treatment. Exploratory factor analysis was used to examine the underlying factor structure of 66 items. Four factors were identified, subscales were named Function, Mood, Feelings, and Awareness, and the 6 items with the highest loadings were included in each scale.11 As the 4 subscales were only slightly correlated, a sum score of the phenomenon could not be derived. Test-retest reliability was, however, found satisfactory for all subscales and construct validity and discriminate ability was acceptable for the subscales Function, Feelings, and Awareness, but responsiveness to change was only acceptable for Function and Awareness.12 As the overall measurement properties of BARQ were not satisfactory and the subscales could not be summarized into a total score, we decided to examine the possibility of developing a unidimensional scale by using Rasch analysis.13 This analysis is increasingly being applied in the development of measurement tools to assess the unidimensionality (the extent to which items measure a single construct, eg, body awareness), item difficulty (the relative difficulty of the items when compared with one another) and person separation (the extent to which items distinguish between distinct levels of the construct, here body awareness). A unidimensional scale would support the assumption that 1 construct is assessed, being a prerequisite for summing all items into a valid total score. If the data fit the model, the raw ordinal level of scores can be transformed into interval level scores. In Rasch analysis items are tested against the expectations of the Rasch measurement model14 and justifies the use of parametric statistics. Such a scale would provide a greater accuracy in scores of importance (ie, when comparing scores between groups).15 The purpose of this study was therefore to explore the possibility of developing a unidimensional and improved version of BARQ by using Rasch analysis. If a revised and improved version of BARQ was developed, a secondary purpose was to investigate its test-retest reliability. Methods The present study was carried out in 3 main steps (Fig. 1). Items originally generated for BARQ development (Tab. 1), and scored by people with musculoskeletal pain11 were included in an initial Rasch analysis (survey 1). Items found to fit the Rasch model were included in a second Rasch analysis (survey 2), performed on data from a new sample of people with long-lasting musculoskeletal pain. The final version of the questionnaire was then assessed for test-retest reliability (survey 3). The different steps of test development are elaborated on below. Informed consent was obtained from all participants. Figure 1. View largeDownload slide Steps in the study, including item generation and 3 surveys. NPMP = Norwegian Psychomotor Physiotherapy. Figure 1. View largeDownload slide Steps in the study, including item generation and 3 surveys. NPMP = Norwegian Psychomotor Physiotherapy. Table 1. Initial Pool of 66 Items Collected for Body Awareness Rating Questionnaire (BARQ) Developmenta Description  Item No.  Location  Fit Residual  Probability  I am stable on my feet  1  0.594  3.222  0  In standing my feet have good contact with the ground  2  0.673  2.347  0  I pull myself together in standing  3  0.327  0.89  .894265  I am not aware of how I am standing  4  −0.054  1.81  .117153  I sense if my joints are flexed or extended  5  0.27  2.273  0  I can sense if my joints are tense or flexible  6  0.614  0.468  .00439  I cannot find comfortable positions in standing  7b  −0.058  −1.467  0  I am often tense  8b,c,d  0.276  −1.449  0  I am not aware of my habit positions  9  0.243  1.668  .003417  I never sit comfortably  10b,d  −0.115  −1.957  0  I can’t get comfortable when I’m lying down  11b,c  −0.197  −1.655  .000561  I search for comfortable positions  12  0.697  −0.6  .338519  My body sensation helps me find comfortable positions  13  0.854  0.663  .041528  I am always holding my breath  14b  −0.264  −1.642  .000096  I am not aware of the way I breathe  15b,c,e  0.106  0.29  .670769  I breathe more easily when I’m in a good mood  16f  −0.621  1.872  0  I try to control my breath  17b  −0.197  −0.458  .001362  I try to breathe correctly  18  0.198  1.3  .177791  I am moving smoothly  19  −0.137  4.945  0  When I am aware that I am holding breath I relax  20  0.44  0.763  .017304  My body is functioning well on a daily basis  21  −0.02  6.133  0  I always hurry even when I don’t have to  22e  0.035  0.375  .72075  I avoid being exhausted if I do not have to  23  −0.184  0.258  .6167  I am fond of moving  24  0.496  2.18  0  I don’t pay attention to the way I move  25b,c,e  −0.117  0.986  .600102  I have problems moving my arms and legs  26b  −0.381  −1.329  .001126  I move to improve my mood  27  0.214  1.867  .004535  I am afraid of moving  28b,g  −0.689  −0.913  .005582  My body is not understandable  29  −0.091  −2.132  0  My body is unpredictable  30c,d  −0.055  −1.794  0  My body is tense without me knowing why  31b,c  0.1  −2.266  0  I am tense in social contexts  32b  0.04  −1.278  .005275  I struggle to relax  33b,c  0.203  −2.342  0  My hands are cold and wet when I feel uncertain  34  −0.26  0.142  .816596  I relax more easily when I’m feeling safe  35f  −0.477  1.292  .000271  My muscles are often tense  36b,d  0.508  −1.886  0  I am able to sense if my muscles are tense  37  0.348  1.712  .011598  I am able to relax when I am feeling tense  38  −0.211  3.781  0  I don’t like to be touched  39b,c,g  −0.61  −0.721  .189535  I like being touched in natural situations  40  0.467  1.163  .000001  My heart rate is high even though I am not exhausted  41b  −0.229  −1.487  .001797  I feel dizzy not for any reason  42  −0.147  −1.85  .007996  I feel sick to my stomach, when faced with what I dislike  43g  −0.225  0.198  .525221  My digestion is affected by how I feel  44c,f  −0.058  4.567  0  My body always hurts  45d  −0.3  −2.683  0  I am able to sense body parts  46  −0.152  0.228  .663886  I do not understand why I have pain  47b  0.065  −1.097  .006182  My pain is affected by my mood  48f  −0.138  3.313  0  I can relieve my pain  49  −0.133  3.649  0  My body is affected by how I feel  50c,f  −0.033  1.89  .001665  My body feels different when I am happy and when I am sad  51f  −0.342  3.225  0  I try not to show how I’m feeling  52b,c  −0.212  −0.703  .039561  My body feels like an alien  53b  0.143  −1.646  0  I dislike my body  54b,g  −0.354  −2.12  0  I tense up in order to hide my feelings  55  −0.026  −2.295  0  I am ashamed of my body  56b,g  −0.458  −1.64  .000143  I am ashamed of certain parts of my body  57g  −0.212  −1.562  .003217  My feelings are expressed through my body  58  0.077  1.161  .016487  I avoid sensing my body  59b,c  −0.388  −1.607  .000019  I don’t realize that I am tired until I am exhausted  60b,e  −0.056  −1.381  .00215  It is not important to consider bodily needs  61b,e  −0.562  0.12  .479489  I know how to find new strength  62  0.137  4.717  0  I take a break when I feel tired  63  0.187  3.472  0  I always push myself to my limits  64b,e  0.066  −0.967  .000616  Body signals help me to find my limits  65  0.23  4.096  0  I do everything to manage my life on a daily basis, but still I feel bad  66b,d  0.158  −1.985  0  Description  Item No.  Location  Fit Residual  Probability  I am stable on my feet  1  0.594  3.222  0  In standing my feet have good contact with the ground  2  0.673  2.347  0  I pull myself together in standing  3  0.327  0.89  .894265  I am not aware of how I am standing  4  −0.054  1.81  .117153  I sense if my joints are flexed or extended  5  0.27  2.273  0  I can sense if my joints are tense or flexible  6  0.614  0.468  .00439  I cannot find comfortable positions in standing  7b  −0.058  −1.467  0  I am often tense  8b,c,d  0.276  −1.449  0  I am not aware of my habit positions  9  0.243  1.668  .003417  I never sit comfortably  10b,d  −0.115  −1.957  0  I can’t get comfortable when I’m lying down  11b,c  −0.197  −1.655  .000561  I search for comfortable positions  12  0.697  −0.6  .338519  My body sensation helps me find comfortable positions  13  0.854  0.663  .041528  I am always holding my breath  14b  −0.264  −1.642  .000096  I am not aware of the way I breathe  15b,c,e  0.106  0.29  .670769  I breathe more easily when I’m in a good mood  16f  −0.621  1.872  0  I try to control my breath  17b  −0.197  −0.458  .001362  I try to breathe correctly  18  0.198  1.3  .177791  I am moving smoothly  19  −0.137  4.945  0  When I am aware that I am holding breath I relax  20  0.44  0.763  .017304  My body is functioning well on a daily basis  21  −0.02  6.133  0  I always hurry even when I don’t have to  22e  0.035  0.375  .72075  I avoid being exhausted if I do not have to  23  −0.184  0.258  .6167  I am fond of moving  24  0.496  2.18  0  I don’t pay attention to the way I move  25b,c,e  −0.117  0.986  .600102  I have problems moving my arms and legs  26b  −0.381  −1.329  .001126  I move to improve my mood  27  0.214  1.867  .004535  I am afraid of moving  28b,g  −0.689  −0.913  .005582  My body is not understandable  29  −0.091  −2.132  0  My body is unpredictable  30c,d  −0.055  −1.794  0  My body is tense without me knowing why  31b,c  0.1  −2.266  0  I am tense in social contexts  32b  0.04  −1.278  .005275  I struggle to relax  33b,c  0.203  −2.342  0  My hands are cold and wet when I feel uncertain  34  −0.26  0.142  .816596  I relax more easily when I’m feeling safe  35f  −0.477  1.292  .000271  My muscles are often tense  36b,d  0.508  −1.886  0  I am able to sense if my muscles are tense  37  0.348  1.712  .011598  I am able to relax when I am feeling tense  38  −0.211  3.781  0  I don’t like to be touched  39b,c,g  −0.61  −0.721  .189535  I like being touched in natural situations  40  0.467  1.163  .000001  My heart rate is high even though I am not exhausted  41b  −0.229  −1.487  .001797  I feel dizzy not for any reason  42  −0.147  −1.85  .007996  I feel sick to my stomach, when faced with what I dislike  43g  −0.225  0.198  .525221  My digestion is affected by how I feel  44c,f  −0.058  4.567  0  My body always hurts  45d  −0.3  −2.683  0  I am able to sense body parts  46  −0.152  0.228  .663886  I do not understand why I have pain  47b  0.065  −1.097  .006182  My pain is affected by my mood  48f  −0.138  3.313  0  I can relieve my pain  49  −0.133  3.649  0  My body is affected by how I feel  50c,f  −0.033  1.89  .001665  My body feels different when I am happy and when I am sad  51f  −0.342  3.225  0  I try not to show how I’m feeling  52b,c  −0.212  −0.703  .039561  My body feels like an alien  53b  0.143  −1.646  0  I dislike my body  54b,g  −0.354  −2.12  0  I tense up in order to hide my feelings  55  −0.026  −2.295  0  I am ashamed of my body  56b,g  −0.458  −1.64  .000143  I am ashamed of certain parts of my body  57g  −0.212  −1.562  .003217  My feelings are expressed through my body  58  0.077  1.161  .016487  I avoid sensing my body  59b,c  −0.388  −1.607  .000019  I don’t realize that I am tired until I am exhausted  60b,e  −0.056  −1.381  .00215  It is not important to consider bodily needs  61b,e  −0.562  0.12  .479489  I know how to find new strength  62  0.137  4.717  0  I take a break when I feel tired  63  0.187  3.472  0  I always push myself to my limits  64b,e  0.066  −0.967  .000616  Body signals help me to find my limits  65  0.23  4.096  0  I do everything to manage my life on a daily basis, but still I feel bad  66b,d  0.158  −1.985  0  aAs presented to the first sample of participants (N = 300) and showing the disposition of the items. bItems remaining after the first Rasch analysis. cItems included in the revised BARQ. dItems included in the original BARQ subscale Function. eItems included in the original BARQ subscale Mood. fItems included in the original BARQ subscale Feelings. gItems included in the original BARQ subscale Awareness. View Large Participants A convenience sampling strategy was applied (Tab. 2). The participants were people who had musculoskeletal and/or psychosomatic complaints who consulted their primary physician for their complaints. After examination the physician referred them to NPMP treatment. Taking the inclusion and exclusion criteria into consideration, 6 NPMP specialists recruited participants consecutively to the study. Five specialists worked in primary health care, and 1 worked in an outpatient hospital clinic. The inclusion criteria were long-lasting (>3 months) musculoskeletal pain and/or psychosomatic disorders and reading skills in Norwegian. Exclusion criteria were psychiatric disorders like severe depression and anxiety. Data was collected for survey 1 in 2008, for survey 2 in 2013, and for survey 3 in 2016. Table 2. Participant Characteristics Characteristic  Survey 1 (N = 300)a  Survey 2 (n = 125)  Survey 3 (n = 48)  Sex: no. (%) of women  231 (77)  91 (73)  45 (94)  Age, y, (SD)  42.4 (11.1)  42.1 (11.0)  46.9 (12.0)  Duration of pain, y, (SD)  8.2 (7.8)  8.9 (7.7)  11.4 (7.9)  Type of work (% of participants)  Blue collar  50  43  39.6  White collar  50  57  60.4  Characteristic  Survey 1 (N = 300)a  Survey 2 (n = 125)  Survey 3 (n = 48)  Sex: no. (%) of women  231 (77)  91 (73)  45 (94)  Age, y, (SD)  42.4 (11.1)  42.1 (11.0)  46.9 (12.0)  Duration of pain, y, (SD)  8.2 (7.8)  8.9 (7.7)  11.4 (7.9)  Type of work (% of participants)  Blue collar  50  43  39.6  White collar  50  57  60.4  a14% without musculoskeletal problems. View Large Item Generation The items included in the first Rasch analysis had been collected during an extensive process with focus group interviews including NPMP physical therapists (2 groups) and people with long-lasting musculoskeletal pain (4 groups). The generated items were supplemented and commented on in written by NPMP physical therapists from different parts of Norway. All collected items were then field tested on a sample of 30 people with long-lasting musculoskeletal pain. The total item pool finally included 66 items, all refined into statements to be rated on a 7-point response scale, ranging from 1 (totally agree) to 7 (totally disagree). Twenty-four of these items, 6 in each of 4 subscales, were included in the initial BARQ (further elaborated in Dragesund et al11). First Survey The 66 items (same as above) were scored on 7-point response scales by a sample of 300 people with long-lasting musculoskeletal pain. Rasch analyses were performed to explore the possibility for obtaining a unidimensional scale. Second Survey A second survey was performed to examine whether the results from the first Rasch analysis could be confirmed in a new sample of people (n = 127). For the sake of content coverage, we included all items supported by the first Rasch analysis, plus items from the original 4 subscales of BARQ that were not supported by the first Rasch analysis, as well as a few items that were considered important from the original pool of 66 items. Third Survey Test-retest reliability was based on data from a new sample of 48 people with long-lasting musculoskeletal pain. These participants completed the revised version of the BARQ from step 2, twice within 1 week. Data Analysis The demographic and item scores were entered into SPSS (versions 17 and 21; SPSS, Chicago, Illinois) and RUMM2020 and RUMM2030 software packages (RUMM Laboratory; www.rummlab.com). The Rasch model assumes that the probability of a person affirming a trait, like in this case body awareness, in an item of a measure depends on the person's level of that trait (θ) and the level of body awareness required by the item (b). In a model in which there are several response categories, the model is expressed in a formula.16 ln (Pni/1-Pni) = θn-bi-τi. Pni is the probability that a person n will affirm the item, θn is the person's level of the trait, bi is the level of the trait expressed by the item, and τi represents the .5 probability point (threshold) between adjacent response levels for that item. In this case it is assumed that as a person's body awareness increases, the probability of a maximum score on the item decreases. Person and item scores are used to “calibrate” items on a logit scale. Items at one end of the scale are “easier,” whereas items at the other end are “more difficult.” In the present analyses, items with a positive calibration were more difficult (ie, indicated less body awareness). A logit is the natural log-odds of the level of difficulty of a particular item in relation to all other items in the scale (hierarchy). Rasch analysis also constructs a hierarchy of the respondents; the hierarchy was ordered by the respondents’ level of disability. Unidimensionality refers to the single underlying construct measured by items that form a scale; for this instrument body awareness. The adequacy of the fit of each item to the Rasch model is assessed by the overall model fit, individual person fit and item fit, thresholds of the response categories, local dependency, differential item functioning (DIF), and unidimensionality. This requires a number of tests for effecting a transformation of the ordinal, raw scores into linear, interval level measures.17,18 The overall fit of data to the model was examined with item-trait interaction statistics. A nonsignificant chi-square probability value, Bonferroni adjusted, indicates the hierarchical ordering of the items to be consistent across all levels of the trait, body awareness. The fit of individual items and persons was examined by inspecting their fit residuals, the difference between their observed and expected logit values. If the residuals were beyond the standardized ±2.5 (99% CI) or if an item showed a significant chi-square probability value, misfit was considered. The mean of all fit residual across all items of the scale and across all person estimates should be closed to 0 with a SD close to 1 (preferably < 1.40). The local response independency was considered to be dependent if the residual correlation was greater than .2 above the average of all residual correlations.19 Unidimensionality The concept of unidimensionality is a primary tenet underpinning the Rasch analysis. If the scores on the revised version of the BARQ are to be added into a total score representing an overall level of body awareness, the analysis should indicate a unidimensional scale, measuring 1 construct. The unidimensionality was tested by contrasting items loading positively and negatively on the first principle component of the residuals, through series of independent t tests.20 The present of significant multidimensionality was noted if more than 5% of these t tests were significant or if the lower CI of the observed proportion was below 0.05.21 Response threshold ordering In Rasch analysis the response options for each item should reflect a logical progression of the underlying trait, body awareness. If this is not demonstrated, thresholds are disordered and can be re-scored by collapsing categories.15 DIF To explore potential bias across groups of respondents, analysis for DIF was performed for 3 factors: sex, age (2 groups: ≥ 43 years and < 43 years), and duration of pain (2 groups: ≥ 5 years and <5 years). Rasch analysis generates a Person Separation Index, which is a reliability value equivalent to the Cronbach alpha for internal consistency. A Person Separation Index of greater than 0.70 was considered desirable for comparing groups and also as evidence of sufficient discriminative ability.22 Targeting of Persons and Items In clinical practice it is important that the items are appropriately targeted for the population for whom the questionnaire is developed. Poorly targeted questionnaires result in floor and ceiling effects. Scale-to-sample targeting was made by comparing person locations from the sample with locations of the scale, requiring means close to 0 and an SD close to 1.15 This might also be visualized by a person-item threshold distribution histogram, which informs about the suitability of the sample for evaluating the scale and the suitability of the scale for measuring the sample. Test-Retest Reliability The COnsensus-based Standards for the selection of health status Measurement InstrumeNts (COSMIN) guidelines were followed.23 The intraclass correlation coefficient [ICC(2.1)] was used to examine test-retest reliability.24 An ICC of .80 or more was considered evidence of satisfactory test-retest reliability. Repeatability was assessed by calculating the standard error of measurement between scores at test and retest, being the square root of the mean square error (total in the analysis-of-variance table). The 95% CI for repeatability was further calculated by multiplying the standard error of measurement by 2.77.25 This value is often referred to as the minimal detectable change, being the smallest change that must be observed between 2 measurements before it can be considered to exceed measurement error with 95% certainty. Role of the Funding Source The Norwegian Fund for Postgraduate Training in Physiotherapy provided financial support of the study. The funding source had no role in the study. Results Characteristics of Participants Characteristics of the participants taking part in the 3 surveys are presented in Table 2. The first survey included 300 participants, mostly women (n = 231). The mean age was 42.4 years. Most participants (86%) were treated for long-lasting musculoskeletal and psychosomatic disorders, with an average duration of pain of 8.2 years. The remaining participants (14%), being healthy without musculoskeletal pain problems. In the second survey there were 125 participants, also mostly women (n = 91). The mean age was 42.1 years. They all had long-lasting musculoskeletal pain, with average duration of pain of 8.9 years. The third survey included 48 (93.8% women), all with long-lasting musculoskeletal pain. The mean age was 46.9 years, and their mean duration of pain was 11.4 years. Survey 1: First Rasch Analysis of Initial Item Pool All 66 items were scored on a 7-point scale by 300 people in the first survey (Tab. 1). The missing data on the initial 66 items ranged from 0.3% to 11%. For Rasch analysis of polytomous scales there are 2 models to be used: the unrestricted model or the rating scale model. The unrestricted model places no constraints on the threshold parameters, whereas the rating scale model constrains all thresholds to be equally spaced across the trait of the items. Results on the likelihood ratio test decide which model to choose. The likelihood ratio test was significant (P < .0000001), suggesting that the unrestricted model should be adopted. The overall statistics indicated that the 66 items did not reflect a unidimensional measure (P = .00076). The analysis showed misfit of both persons and items outside the fit residual range of ±2.5. This might have been due to the fact that all the 66 items had disordered thresholds. All items had to be rescored to a 3-point response scale in order to avoid disordered thresholds (scores of 1 and 2 = 0, scores of 3 to 5 = 1, and scores of 6 and 7 = 2). Only 4 of the items (40, 42, 43, and 57) had disordered thresholds after the rescoring, suggesting that the 3-point scale worked much better than the 7-point scale. The overall statistics were still highly significant (P = .000). Fifty-two participants had extreme scores outside the ±2.5 limit, 1 item (55) was redundant (fit residual = −2.63), and 10 items (19, 21, 38, 44, 48, 49, 51, 62, 63, and 65) had fit residuals above 2.5. In addition, several of the items had poor statistics in terms of a significant chi-square probability value (adjusted for Bonferroni: .00076) (items 29, 54, 33, 45, 31, 10, 36, 7, 8, 40, 16, 2, 5, 9, 24, and 1). Removal of misfitting and redundant items were carried out during 11 following analyses. The process ended up with a unidimensional 26-item version (Tab. 3). Thirteen of these 26 items overlapped with items from the original BARQ, while eleven were excluded during the analysis. When checking the DIF of this version, a DIF for age group was found for items 8 and 56, and a DIF for duration of pain was found for items 15, 36, and 66. There was no DIF for sex. Table 3. Twenty-six Items Resulting From First Rasch Analysisa Item Description  Item No.b  Item Difficulty (Logits)  Standard Error for Location  Fit Residual  Probability  My muscles are often tense  36  2.269  0.508  −1.886  0  I am often tense  8  1.273  0.276  −1.449  0  I do everything to manage my life on a daily basis, but still I feel bad  66  1.137  0.158  −1.985  0  I struggle to relax  33  0.955  0.203  −2.342  0  I try not to show my feelings  52  0.763  0.143  −0.703  .039561  My body feels tense and I do not know why  31  0.753  0.1  −2.266  0  I always push myself to my limits  64  0.606  0.066  4.096  0  I am not aware of the way I breathe  15  0.489  0.106  0.29  .670769  I am tense in social contexts  32  0.435  0.04  −1.278  .005275  I try to control my breath  17  0.142  −0.197  −0.458  .001362  I cannot find comfortable positions in standing  7  0.13  −0.058  −1.467  0  I don’t realize that I am tired until I am exhausted  60  0.086  −0.056  −1.381  .002150  I never sit comfortably  10  0.026  −0.115  −1.957  0  I can’t get comfortable when I’m lying down  11  −0.077  −0.197  −1.655  .000561  I do not understand why I have pain  47  −0.168  −0.138  −1.097  .006182  I am always holding my breath  14  −0.212  −0.264  −1.642  .000096  I don’t pay attention to the way I move  25  −0.217  −0.117  0.986  .600102  My heart rate is high even though I am not exhausted  41  −0.379  −0.229  −1.487  .001797  I dislike my body  54  −0.474  −0.3  −2.12  0  I avoid sensing my body  59  −0.567  −0.388  −1.607  .000019  I have problems moving my arms and legs  26  −0.69  −0.381  −1.329  .001126  My body feels like an alien  53  −0.771  −0.354  −1.646  0  I am ashamed of my body  56  −1.071  −0.458  −1.64  .000143  It is not important to consider bodily needs  61  −1.357  −0.562  0.12  .479489  I don’t like to be touched  39  −1.467  −0.61  −0.721  .189535  I am afraid of moving  28  −1.615  −0.689  −0.913  .005582  Item Description  Item No.b  Item Difficulty (Logits)  Standard Error for Location  Fit Residual  Probability  My muscles are often tense  36  2.269  0.508  −1.886  0  I am often tense  8  1.273  0.276  −1.449  0  I do everything to manage my life on a daily basis, but still I feel bad  66  1.137  0.158  −1.985  0  I struggle to relax  33  0.955  0.203  −2.342  0  I try not to show my feelings  52  0.763  0.143  −0.703  .039561  My body feels tense and I do not know why  31  0.753  0.1  −2.266  0  I always push myself to my limits  64  0.606  0.066  4.096  0  I am not aware of the way I breathe  15  0.489  0.106  0.29  .670769  I am tense in social contexts  32  0.435  0.04  −1.278  .005275  I try to control my breath  17  0.142  −0.197  −0.458  .001362  I cannot find comfortable positions in standing  7  0.13  −0.058  −1.467  0  I don’t realize that I am tired until I am exhausted  60  0.086  −0.056  −1.381  .002150  I never sit comfortably  10  0.026  −0.115  −1.957  0  I can’t get comfortable when I’m lying down  11  −0.077  −0.197  −1.655  .000561  I do not understand why I have pain  47  −0.168  −0.138  −1.097  .006182  I am always holding my breath  14  −0.212  −0.264  −1.642  .000096  I don’t pay attention to the way I move  25  −0.217  −0.117  0.986  .600102  My heart rate is high even though I am not exhausted  41  −0.379  −0.229  −1.487  .001797  I dislike my body  54  −0.474  −0.3  −2.12  0  I avoid sensing my body  59  −0.567  −0.388  −1.607  .000019  I have problems moving my arms and legs  26  −0.69  −0.381  −1.329  .001126  My body feels like an alien  53  −0.771  −0.354  −1.646  0  I am ashamed of my body  56  −1.071  −0.458  −1.64  .000143  It is not important to consider bodily needs  61  −1.357  −0.562  0.12  .479489  I don’t like to be touched  39  −1.467  −0.61  −0.721  .189535  I am afraid of moving  28  −1.615  −0.689  −0.913  .005582  aThe items are presented in location order, with easy items on the top and difficult items on the bottom. bFrom the initial pool of 66 items. View Large Survey 2: Second Rasch Analysis The second survey included 39 items that were scored on a 4-point response scale, from 0 (totally agree) to 3 (totally disagree); 26 items resulting from the first Rasch analysis, 11 items from the original BARQ subscales that were excluded in the first Rasch analysis, and 2 items from the original items pool that seemed clinically important. A 4-point response scale was chosen in order to improve the reliability of the scale. The missing data on these 39 items were low and ranged from 0.3% to 3.1%. The unrestricted model was initially adopted (P < .0000001) on all 127 people. The overall statistics were significant (P = .036), indicating that the initial 39 items did not reflect a unidimensional measure. The residual mean value for persons (−0.03; SD = 1.51), indicated some misfit among respondents in the sample, and individual item fit, adjusted for Bonferroni, also indicated misfit of items 40 and 62. Two participants were excluded from the analysis because they had not reported sex or age categories. Scores of the remaining 125 participants were thus included in the further analysis. Item 40 and item 62 were deleted because of fit residual of greater than 2.5. Items with local independency (items 32, 36, and 47), items with similar locations (reflecting degree of difficulty) (items 7, 14, 16, 17, 22, 26, 28, 35, 41, 43, 45, 48, 51, 53, 54, 56, 57, 60, 61, 64, and 66), as well as 1 respondent not responding as expected (id no. 113) were systematically deleted in order to improve the data fit to the Rasch model and to achieve a questionnaire with fewer items (Tab. 4). The process ended up with a 12-item solution with a 4-point response scale (Tab. 5). The Norwegian and English versions of the final revised version of the BARQ (BARQ-R) are presented in eAppendixes 1 and 2. Table 4. Thirty-nine Items Scored by 127 Participants and Included in Second Rasch Analysis Item Description  Item No.a  Item Difficulty (Logits)  Standard Error for Location  Fit Residual  Probability  Reason for Exclusion  I cannot find comfortable positions in standing  7  −0.254  0.108  1.365  .398352  Location similar to other  I am often tense  8  −1.314  0.132  0.572  .990277    I never sit comfortably  10  −0.215  0.103  3.480  .001004  Misfit > 2.5  I can’t get comfortable when I’m lying down  11  0.339  0.105  2.010  .770461    I am always holding my breath  14  0.281  0.111  0.029  .695983  Location similar to other  I am not aware of the way I breathe  15  0.433  0.124  0.325  .674593    I breathe more easily when I’m in a good mood  16  −1.123  0.110  0.596  .617333  Location similar to other  I try to control my breath  17  −0.052  0.114  1.419  .096183  Location similar to other  I always hurry even when I don’t have to  22  −0.463  0.098  −0.623  .720634  Location similar to other  I don’t pay attention to the way I move  25  0.307  0.115  0.535  .399475    I have problems moving my arms and legs  26  0.967  0.121  −0.648  .157807  Location similar to other  I am afraid of moving  28  0.834  0.117  0.738  .213278  Location similar to other  My body is unpredictable  30  0.571  0.108  0.282  .284172    My body is tense without me knowing why  31  −0.133  0.107  −0.275  .143670    I am tense in social contexts  32  0.180  0.104  −2.239  .006189  Local dependency  I struggle to relax  33  −0.305  0.109  −0.567  .774623    I relax more easily when I’m feeling safe  35  0.780  0.099  1.045  .393768  Location similar to other  My muscles are often tense  36  −1.020  0.115  0.264  .765420  Local dependency  I don’t like to be touched  39  1.314  0.129  0.340  .838816    I like being touched in natural situations  40  0.490  0.100  2.923  .000072  Misfit > 2.5  My heart rate is high even though I am not exhausted  41  0.683  0.109  1.189  .032264  Location similar to other  I feel sick to my stomach, when faced with what I dislike  43  0.746  0.111  0.352  .746936  Location similar to other  My digestion is affected by how I feel  44  0.050  0.097  0.357  .545231    My body always hurts  45  −0.082  0.109  −0.651  .370472  Location similar to other  I do not understand why I have pain  47  0.567  0.116  −0.394  .542590  Local dependency  My pain is affected by my mood  48  −0.661  0.106  1.431  .552783  Location similar to other  My body is affected by how I feel  50  −0.710  0.111  1.121  .417151    My body feels different when I am happy and when I am sad  51  −0.524  0.102  −0.007  .339661  Location similar to other  I try not to show how I’m feeling  52  −0.003  0.100  −0.123  .965726    My body feels like an alien  53  0.944  0.115  −0.029  .493674  Location similar to other  I dislike my body  54  0.458  0.109  −0.961  .187169  Location similar to other  I am ashamed of my body  56  1.002  0.116  −0.225  .233640  Location similar to other  I am ashamed of certain parts of my body  57  0.328  0.099  0.210  .328811  Location similar to other  I avoid sensing my body  59  0.931  0.125  −0.017  .915166    I don’t realize that I am tired until I am exhausted  60  0.117  0.105  0.594  .104907  Location similar to other  It is not important to consider bodily needs  61  1.012  0.118  0.588  .224231  Location similar to other  I always push myself to my limits  64  0.277  0.104  1.703  .259346  Location similar to other  I do everything to manage my life on a daily basis, but still I feel bad  66  −0599  0.108  0.539  .174838  Location similar to other  I know how to find new strength  62  0.373  0.117  2.553  .001205  Misfit > 2.5  Item Description  Item No.a  Item Difficulty (Logits)  Standard Error for Location  Fit Residual  Probability  Reason for Exclusion  I cannot find comfortable positions in standing  7  −0.254  0.108  1.365  .398352  Location similar to other  I am often tense  8  −1.314  0.132  0.572  .990277    I never sit comfortably  10  −0.215  0.103  3.480  .001004  Misfit > 2.5  I can’t get comfortable when I’m lying down  11  0.339  0.105  2.010  .770461    I am always holding my breath  14  0.281  0.111  0.029  .695983  Location similar to other  I am not aware of the way I breathe  15  0.433  0.124  0.325  .674593    I breathe more easily when I’m in a good mood  16  −1.123  0.110  0.596  .617333  Location similar to other  I try to control my breath  17  −0.052  0.114  1.419  .096183  Location similar to other  I always hurry even when I don’t have to  22  −0.463  0.098  −0.623  .720634  Location similar to other  I don’t pay attention to the way I move  25  0.307  0.115  0.535  .399475    I have problems moving my arms and legs  26  0.967  0.121  −0.648  .157807  Location similar to other  I am afraid of moving  28  0.834  0.117  0.738  .213278  Location similar to other  My body is unpredictable  30  0.571  0.108  0.282  .284172    My body is tense without me knowing why  31  −0.133  0.107  −0.275  .143670    I am tense in social contexts  32  0.180  0.104  −2.239  .006189  Local dependency  I struggle to relax  33  −0.305  0.109  −0.567  .774623    I relax more easily when I’m feeling safe  35  0.780  0.099  1.045  .393768  Location similar to other  My muscles are often tense  36  −1.020  0.115  0.264  .765420  Local dependency  I don’t like to be touched  39  1.314  0.129  0.340  .838816    I like being touched in natural situations  40  0.490  0.100  2.923  .000072  Misfit > 2.5  My heart rate is high even though I am not exhausted  41  0.683  0.109  1.189  .032264  Location similar to other  I feel sick to my stomach, when faced with what I dislike  43  0.746  0.111  0.352  .746936  Location similar to other  My digestion is affected by how I feel  44  0.050  0.097  0.357  .545231    My body always hurts  45  −0.082  0.109  −0.651  .370472  Location similar to other  I do not understand why I have pain  47  0.567  0.116  −0.394  .542590  Local dependency  My pain is affected by my mood  48  −0.661  0.106  1.431  .552783  Location similar to other  My body is affected by how I feel  50  −0.710  0.111  1.121  .417151    My body feels different when I am happy and when I am sad  51  −0.524  0.102  −0.007  .339661  Location similar to other  I try not to show how I’m feeling  52  −0.003  0.100  −0.123  .965726    My body feels like an alien  53  0.944  0.115  −0.029  .493674  Location similar to other  I dislike my body  54  0.458  0.109  −0.961  .187169  Location similar to other  I am ashamed of my body  56  1.002  0.116  −0.225  .233640  Location similar to other  I am ashamed of certain parts of my body  57  0.328  0.099  0.210  .328811  Location similar to other  I avoid sensing my body  59  0.931  0.125  −0.017  .915166    I don’t realize that I am tired until I am exhausted  60  0.117  0.105  0.594  .104907  Location similar to other  It is not important to consider bodily needs  61  1.012  0.118  0.588  .224231  Location similar to other  I always push myself to my limits  64  0.277  0.104  1.703  .259346  Location similar to other  I do everything to manage my life on a daily basis, but still I feel bad  66  −0599  0.108  0.539  .174838  Location similar to other  I know how to find new strength  62  0.373  0.117  2.553  .001205  Misfit > 2.5  aFrom the initial pool of 66 items. View Large Table 5. Final Revised Body Awareness Rating Questionnaire, With 4-Point Response Scalea Item Description  Completely 
Disagree (0)  Somewhat 
Disagree (1)  Somewhat 
Agree (2)  Completely 
Agree (3)  1. I am often tenseb          2. My body is affected by how I feelb          3. I am not aware of the way I breatheb          4. I don’t pay attention to the way I moveb          5. I struggle to relax          6. My body is tense without me knowing why          7. I try not to show how I’m feeling          8. My digestion is affected by how I feelb          9. I can’t get comfortable when I’m lying down          10. My body is unpredictableb          11. I avoid paying too much attention to my body          12. I don’t like to be touchedb          Item Description  Completely 
Disagree (0)  Somewhat 
Disagree (1)  Somewhat 
Agree (2)  Completely 
Agree (3)  1. I am often tenseb          2. My body is affected by how I feelb          3. I am not aware of the way I breatheb          4. I don’t pay attention to the way I moveb          5. I struggle to relax          6. My body is tense without me knowing why          7. I try not to show how I’m feeling          8. My digestion is affected by how I feelb          9. I can’t get comfortable when I’m lying down          10. My body is unpredictableb          11. I avoid paying too much attention to my body          12. I don’t like to be touchedb          aThe items are presented in location order, with easier items on the top and more difficult items on the bottom. bAlso included in the original Body Awareness Rating Questionnaire. View Large The final 12-item solution showed a good fit to the Rasch model, with a chi-square value of 13.63 and a P value of .954 (Bonferroni adjusted 0.5/12 = .004). All the items had satisfactory fit residual, in the range of −2.010 to 0.567 and the item difficulty levels ranged from 1.314 (most difficult items) to −1.314 (least difficult items). The Person Separation Index was 0.76. None of the 12 items had disordered threshold. Furthermore, no significant DIF was demonstrated by sex, age groups and duration of pain, making further adjustment unnecessary. And finally the independent t test between the 2 subsets of items with positive and negative loadings on the first residual component was not statistically significant (P = .159), supporting the unidimensionality of the 12 items. Among these 12 items, 7 items were included in the 4 subscales of the original BARQ (items 8, 50, 30, 44, 39, 15, 25, and 2). Additionally, 9 of the 12 items (8, 15, 25, 33, 31, 52, 11, 59, and 39) were confirmed from the result of the first Rasch analysis (Tab. 1). The 12 items also seem to target the population for whom they are developed (mean person location = −0.04; SD = 0.74) (eFigure; available at https://academic.oup.com/ptj). Survey 3: Test-Retest Reliability The mean scores on test and retest were 20.6 (SD = 5.5) and 20.2 (SD = 5.4), respectively. The mean difference between test and retest for the 12-item scale (sum score ranging from 0 to 36) was 0.44 (SD = 3.2), and the ICC was .83 (95% CI = .71–.89). The standard error of measurement was 2.26, and the smallest detectable change with 95% CI was 6.26, or 17.4% of the total score of 36. This implies that a change in individual participants should be above this value to indicate a change above measurement error. Discussion In the present study, we succeeded in developing a unidimensional assessment tool of body awareness based on previously collected test items for BARQ by using Rasch analysis. The BARQ-R includes 12 self-reported items scored on a 4-point ordinal scale (0–3), reflecting degrees of body awareness even though the items included in this solution is slightly different from the 4-factor solution (original BARQ). The scale also seems to express different levels of problems with body awareness as seen in the clinic, from the most common problems (feeling tense and be affected by mood) to the most challenging (feeling that the body is unpredictable and not liking to be touched). The results from the study indicate that the BARQ-R is a valid and reliable assessment tool that can be used to capture body awareness as reported by people with musculoskeletal pain. For comparison the original BARQ included 24 items in 4 subscales (6 items in each subscale) that could not be combined into a sum score. The BARQ-R is briefer and more user-friendly, taking only 3 to 5 minutes to complete. We therefore recommend applying the BARQ-R for further research and clinical use. The BARQ-R includes a variety of items concerning different aspects of body awareness, like awareness of muscle tension, breath, movements, and the experience of coherence between body and emotions.2 Many physical therapists specialized in a body-mind approach of physical therapy (NPMP) contributed to the generation of items, as well as a large target group of people with musculoskeletal and/or psychosomatic problems. Physical therapists and patients might perceive the phenomenon of body awareness differently, and commonly clinical measurements are based on clinician's opinion only and therefore lack content validity from the perspective of patients. Thus, a major strength of the BARQ-R as well as of the original BARQ is content validity with items derived from both the physical therapists’ and the patients’ perspectives. This is in line with quality criteria recommended by the COSMIN group, underscoring the need to include both patients and therapists when constructing a questionnaire.26 The demographic characteristics of the sample were accordingly quite similar to those of patients on waiting-lists for NPMP described by Breitve et al.27 In the present study Rasch analysis was used to develop a unidimensional measure, which is considered an essential quality when summing up individual items to obtain a valid total score.13,14 The Rasch model enables analysis of targeting of items’ difficulty to persons’ ability by calibrating items and persons on a common scale with interval level units (logits). As such, a primary strength of Rasch analysis is that ordinal scores may be converted into interval level scores. Furthermore the method makes it possible to analyze whether items exhibit a consistent and invariant hierarchy of difficulty,13,14 and if response options for each item reflect a logical progression of the underlying trait.15 The first Rasch analyses clearly showed that the original 7-point response scale did not function for any of the 66 items. Threshold disordering was avoided by collapsing them into a 3-point response scale. This aspect of scoring alternatives of the response scale was not examined in the original BARQ, which was developed based on classical test theory. A scale with disordered threshold will be problematic when used to evaluate effect of an intervention, because it will be difficult to evaluate a change in categories.28 Since it was obvious that the participants were not able to distinguish between 7 levels of body awareness (7-point response scale), we used a 3-point scale in the first Rasch analysis by collapsing scoring alternatives, and found it to be appropriate. However, as we considered a 4-point response scale to be more informative and sensitive to differences in the trait than a 3-point scale, we tried out a 4-point scale when new data were collected in the second Rasch analysis, showing satisfactory results. This is in line with Streiner and Norman,22 who stated that an increase in the number of response options on a rating scale will increase the information value of any observation and improve the reliability, although the agreement will decrease. The 4-point response scale in the second Rasch analysis was shown to improve the reliability of the scale as compared to the results of the first Rasch analysis with a 3-point response scale. Nine of 26 items from the first Rasch analysis (based on 66 items) were confirmed in the second Rasch analysis (based on 39 items), and 7 items of the final 12-item scale (BARQ-R), represented all 4 subscales of the original BARQ (2 from the subscale Function, 2 from the subscale Mood, 1 from the subscale Feelings, and 2 from the subscale Awareness) (Tabs. 1 and 5). The items are also ranked hierarchically according to difficulties of the phenomenon of body awareness across age groups, sexes and duration of pain. The evidence of test-retest reliability was also confirmed, giving an ICC of greater than .80. The smallest detectable change, however, was high, being 6.26, or 17.4% of the total score. Still, the BARQ-R has undergone an extensive improvement in measurement properties. Limitations of the current study concern the sample size in the second survey. Linacre29 has, when using Rasch analysis, proposed a sample size of 250 in order to achieve appropriate degree of precision. The first Rasch analysis included a group of 300 target people, in line with these recommendations. In the second Rasch analysis, however, the sample size was only 125, below the recommended number. To conclude, the present study provides support for a satisfactory internal construct validity and reliability of the 12-item BARQ-R. Thus, the BARQ-R can be recommended used by physical therapists focusing on body awareness among people with long-lasting musculoskeletal pain. Further research to assess other aspects of validity and responsiveness should be carried out. Author Contributions and Acknowledgments Concept/idea/research design: L.I. Strand, M. Grotle Writing: T. Dragesund, L.I. Strand, M. Grotle Data collection: T. Dragesund Data analysis: T. Dragesund, M. Grotle Project management: T. Dragesund Providing facilities/equipment: L.I. Strand Providing institutional liaisons: L.I. Strand Consultation (including review of manuscript before submitting): L.I. Strand, M. Grotle The authors thank the Norwegian Fund for Postgraduate Training in Physiotherapy for financial support of the study. They also thank the physical therapists who recruited participants for the study as well as the participants. Ethics Approval The study was approved by the regional ethics committee (REK West 2012/2238). References 1 Kvåle A, Ljunggren AE. Body awareness therapies. In: Schmidt RF, Willis WD, eds. Encyclopedia of Pain . Berlin, Germany: Springer Verlag; 2007: 167– 169. 2 Thornquist E, Bunkan BH. What Is Psychomotor Physiotherapy?  Oslo, Norway: Norwegian University Press; 1991. 3 Øien AM, Råheim M, Iversen S, Steihaug S. Self perception as embodied knowledge: changing processes for patients with chronic pain. Adv Physiother . 2009; 11: 121– 129. Google Scholar CrossRef Search ADS   4 Dragesund T, Raheim M. Norwegian psychomotor physiotherapy and patients with chronic pain: patients’ perspective on body awareness. Physiother Theory Pract . 2008; 24: 243– 254. Google Scholar CrossRef Search ADS PubMed  5 Roxendal G. Body Awareness Therapy and the Body Awareness Scale: Treatment and Evaluation in Physiotherapy [dissertation] . Gothenborg, Sweden: University of Gothenborg; 1985. 6 Skjærven LH, Kristoffersen K, Gard G. How can movement quality be promoted in clinical practice? A phenomenological study of physical therapist experts. Phys Ther . 2010; 90: 1– 15. Google Scholar CrossRef Search ADS   7 Bruchon-Schweitzer M. Dimensionality of the body-image: the Body-Image Questionnaire. Percept Mot Skills . 1987; 65: 887– 892. Google Scholar CrossRef Search ADS   8 Slade PD, Dewey ME, Newton T, Brodie D, Kiemle G. Development and preliminary validation of the Body Satisfaction Scale (BSS). Psychol Health . 1990; 4: 213– 220. Google Scholar CrossRef Search ADS   9 Rosen JC, Srebnik D, Saltzberg E, Wendt S. Development of a body image avoidance questionnaire. Psychol Assess . 1991; 3: 32– 37. Google Scholar CrossRef Search ADS   10 Cash TF, Szymanski ML. The development and validation of the Body-Image Ideals Questionnaire. J Pers Assess . 1995; 64: 466– 477. Google Scholar CrossRef Search ADS PubMed  11 Dragesund T, Ljunggren AE, Kvåle A, Strand LI. Body Awareness Rating Questionnaire: development of a self-administered questionnaire for patients with long-lasting musculoskeletal and psychosomatic disorders. Adv Physiother . 2010; 12: 87– 94. Google Scholar CrossRef Search ADS   12 Dragesund T, Målfrid R, Strand LI. Body Awareness Rating Questionnaire: measurement properties. Physiother Theory Pract . 2012; 28: 515– 528. Google Scholar CrossRef Search ADS PubMed  13 Rasch G. Probabilistic Models for Some Intelligence and Attainment Tests . Chicago, IL: MESA Press; 1980. 14 Karabatsos G. The Rasch model, additive conjoint measurement, and new models of probabilistic measurement theory. J Appl Meas . 2001; 2: 389– 423. Google Scholar PubMed  15 Tennant A, Conaghan PG. The Rasch measurement model in rheumatology: what is it and why use it? When should it be applied, and what should one look for in a Rasch paper? Arthritis Rheum . 2007; 57: 1358– 1362. Google Scholar CrossRef Search ADS PubMed  16 Tennant A, McKenna SP, Hagell P. Application of Rasch analysis in the development and application of quality of life instruments. Value Health . 2004; 7: 22– 26. Google Scholar CrossRef Search ADS PubMed  17 Pallant JF, Tennant A. An introduction to the Rasch measurement model: an example using the Hospital Anxiety and Depression Scale (HADS). Br J Clin Psychol . 2007; 46: 1– 18. Google Scholar CrossRef Search ADS PubMed  18 Grimby G, Tennant A, Tesio L. The use of raw scores from ordinal scales: time to end malpractice? J Rehabil Med . 2012; 44: 97– 98. Google Scholar CrossRef Search ADS PubMed  19 Marais I, Andrich D. Formalizing dimension and response violations of local independence in the unidimensional Rasch model. J Appl Meas . 2008; 9: 200– 215. Google Scholar PubMed  20 Smith EVJr. Detecting and evaluating the impact of multidimensionality using item fit statistics and principal component analysis of residuals. J Appl Meas . 2002; 3: 205– 231. Google Scholar PubMed  21 Tennant A, Pallant JF. Unidimensionality Matters! (A Tale of Two Smiths?). Rasch Meas . 2006; 20: 1048– 1051. 22 Streiner DL, Norman GR. Health Measurement Scales: A Practical Guide to Their Development and Use . Oxford, England: Oxford Medical Publications; 2008. Google Scholar CrossRef Search ADS   23 Mokkink LB, Terwee CB, Knol DL et al.   The COSMIN checklist for evaluating the methodological quality of studies on measurement properties: a clarification of its content. BMC Med Res Methodol . 2010; 10: 22. Google Scholar CrossRef Search ADS PubMed  24 Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull . 1979; 86: 420– 428. Google Scholar CrossRef Search ADS PubMed  25 Bland JM, Altman DG. Measurement error. BMJ . 1996; 312: 1654. Google Scholar CrossRef Search ADS PubMed  26 Mokkink LB, Terwee CB, Patrick DL et al.   The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes. J Clin Epidemiol . 2010; 63: 737– 745. Google Scholar CrossRef Search ADS PubMed  27 Breitve MH, Hynninen M, Kvåle A. Emosjonelle symptomer og helseplager hos pasienter som oppsøker psykomotorisk fysioterapi [Emotional symptoms and health problems in patients consulting Norwegian psychomotor physiotherapy]. Fysioterapeuten . 2008; 75: 19– 26. 28 Chen WH, Lenderking W, Jin Y, Wyrwich KW, Gelhorn H, Revicki DA. Is Rasch model analysis applicable in small sample size pilot studies for assessing item characteristics? An example using PROMIS pain behavior item bank data. Qual Life Res . 2014; 23: 485– 493. Google Scholar CrossRef Search ADS PubMed  29 Linacre JM. Optimizing rating scale category effectiveness. J Appl Meas . 2002; 3: 85– 106. Google Scholar PubMed  © 2017 American Physical Therapy Association http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Physical Therapy Oxford University Press

The Revised Body Awareness Rating Questionnaire: Development Into a Unidimensional Scale Using Rasch Analysis

Loading next page...
 
/lp/ou_press/the-revised-body-awareness-rating-questionnaire-development-into-a-IXDtCnIs6X
Publisher
Oxford University Press
Copyright
© 2017 American Physical Therapy Association
ISSN
0031-9023
eISSN
1538-6724
D.O.I.
10.1093/ptj/pzx111
Publisher site
See Article on Publisher Site

Abstract

Abstract Background The Body Awareness Rating Questionnaire (BARQ) is a self-report questionnaire aimed at capturing how people with long-lasting musculoskeletal pain reflect on their own body awareness. Methods based on classical test theory were applied to the development of the instrument and resulted in 4 subscales. However, the scales were not correlated, and construct validity might be questioned. Objective The primary purpose of this study was to explore the possibility of developing a unidimensional scale from items initially collected for the BARQ using Rasch analysis. A secondary purpose was to investigate the test-retest reliability of a revised version of the BARQ. Design This was a methodological study. Methods Rasch and reliability analyses were performed for 3 samples of participants with long-lasting musculoskeletal pain. The first Rasch analysis was carried out on 66 items generated for the original BARQ and scored by 300 participants. The items supported by the first analysis were scored by a new group of 127 participants and analyzed in a second Rasch analysis. For the test-retest reliability analysis, 48 participants scored the revised BARQ items twice within 1 week. Results The 2-step Rasch analysis resulted in a unidimensional 12-item revised version of the BARQ with a 4-point response scale (scores from 0 to 36). It showed a good fit to the Rasch model, with acceptable internal consistency, satisfactory fit residuals, and no disordered thresholds. Test-retest reliability was high, with an intraclass correlation coefficient of .83 (95% CI = .71–.89) and a smallest detectable change of 6.3 points. Limitations The small sample size in the second Rasch analysis was a study limitation. Conclusions The revised BARQ is a unidimensional and feasible measurement of body awareness, recommended for use in the context of body-mind physical therapy approaches for musculoskeletal conditions. Body awareness is an overall concept of body consciousness, body management and body experience.1 The phenomenon, concerns the experience of coherence between body and emotions, the ability to sense and being confidential with habitual reactions of the body as well as attitudes toward it. Body awareness seems to be a prerequisite for self-experience.2 The phenomenon as addressed in physical therapy deals with the patient's ability to sense muscle tension and body movements, reflecting emotions and attitudes toward own body. Such ability seems to be a prerequisite for readjusting posture and muscle tension in treatment.2–4 The phenomenon of body awareness is accordingly a core aspect addressed in body–mind treatment approaches like Norwegian Psychomotor Physiotherapy (NPMP)2–4 and Basic Body Awareness Therapy (BBAT)5,6 as well as in other treatment modalities used by physical therapists working in the field of musculoskeletal and mental problems. To properly assess and evaluate body awareness in patients with long-lasting musculoskeletal pain, physical therapists need a reliable and valid measurement tool that is feasible for clinical use. The questionnaires retrieved from search in the literature assessed only related constructs like body image and attitude toward own body,7–10 and no appropriate tool was found available to assess body awareness in the context of physical therapy. Therefore a new self-reported questionnaire was developed, the Body Awareness Rating Questionnaire (BARQ).11,12 The BARQ was aimed to capture body awareness among patients with long-lasting musculoskeletal pain. Items considered to reflect body awareness were collected among NPMP therapists and from descriptions among patients receiving or being on a waiting list for NPMP treatment. Exploratory factor analysis was used to examine the underlying factor structure of 66 items. Four factors were identified, subscales were named Function, Mood, Feelings, and Awareness, and the 6 items with the highest loadings were included in each scale.11 As the 4 subscales were only slightly correlated, a sum score of the phenomenon could not be derived. Test-retest reliability was, however, found satisfactory for all subscales and construct validity and discriminate ability was acceptable for the subscales Function, Feelings, and Awareness, but responsiveness to change was only acceptable for Function and Awareness.12 As the overall measurement properties of BARQ were not satisfactory and the subscales could not be summarized into a total score, we decided to examine the possibility of developing a unidimensional scale by using Rasch analysis.13 This analysis is increasingly being applied in the development of measurement tools to assess the unidimensionality (the extent to which items measure a single construct, eg, body awareness), item difficulty (the relative difficulty of the items when compared with one another) and person separation (the extent to which items distinguish between distinct levels of the construct, here body awareness). A unidimensional scale would support the assumption that 1 construct is assessed, being a prerequisite for summing all items into a valid total score. If the data fit the model, the raw ordinal level of scores can be transformed into interval level scores. In Rasch analysis items are tested against the expectations of the Rasch measurement model14 and justifies the use of parametric statistics. Such a scale would provide a greater accuracy in scores of importance (ie, when comparing scores between groups).15 The purpose of this study was therefore to explore the possibility of developing a unidimensional and improved version of BARQ by using Rasch analysis. If a revised and improved version of BARQ was developed, a secondary purpose was to investigate its test-retest reliability. Methods The present study was carried out in 3 main steps (Fig. 1). Items originally generated for BARQ development (Tab. 1), and scored by people with musculoskeletal pain11 were included in an initial Rasch analysis (survey 1). Items found to fit the Rasch model were included in a second Rasch analysis (survey 2), performed on data from a new sample of people with long-lasting musculoskeletal pain. The final version of the questionnaire was then assessed for test-retest reliability (survey 3). The different steps of test development are elaborated on below. Informed consent was obtained from all participants. Figure 1. View largeDownload slide Steps in the study, including item generation and 3 surveys. NPMP = Norwegian Psychomotor Physiotherapy. Figure 1. View largeDownload slide Steps in the study, including item generation and 3 surveys. NPMP = Norwegian Psychomotor Physiotherapy. Table 1. Initial Pool of 66 Items Collected for Body Awareness Rating Questionnaire (BARQ) Developmenta Description  Item No.  Location  Fit Residual  Probability  I am stable on my feet  1  0.594  3.222  0  In standing my feet have good contact with the ground  2  0.673  2.347  0  I pull myself together in standing  3  0.327  0.89  .894265  I am not aware of how I am standing  4  −0.054  1.81  .117153  I sense if my joints are flexed or extended  5  0.27  2.273  0  I can sense if my joints are tense or flexible  6  0.614  0.468  .00439  I cannot find comfortable positions in standing  7b  −0.058  −1.467  0  I am often tense  8b,c,d  0.276  −1.449  0  I am not aware of my habit positions  9  0.243  1.668  .003417  I never sit comfortably  10b,d  −0.115  −1.957  0  I can’t get comfortable when I’m lying down  11b,c  −0.197  −1.655  .000561  I search for comfortable positions  12  0.697  −0.6  .338519  My body sensation helps me find comfortable positions  13  0.854  0.663  .041528  I am always holding my breath  14b  −0.264  −1.642  .000096  I am not aware of the way I breathe  15b,c,e  0.106  0.29  .670769  I breathe more easily when I’m in a good mood  16f  −0.621  1.872  0  I try to control my breath  17b  −0.197  −0.458  .001362  I try to breathe correctly  18  0.198  1.3  .177791  I am moving smoothly  19  −0.137  4.945  0  When I am aware that I am holding breath I relax  20  0.44  0.763  .017304  My body is functioning well on a daily basis  21  −0.02  6.133  0  I always hurry even when I don’t have to  22e  0.035  0.375  .72075  I avoid being exhausted if I do not have to  23  −0.184  0.258  .6167  I am fond of moving  24  0.496  2.18  0  I don’t pay attention to the way I move  25b,c,e  −0.117  0.986  .600102  I have problems moving my arms and legs  26b  −0.381  −1.329  .001126  I move to improve my mood  27  0.214  1.867  .004535  I am afraid of moving  28b,g  −0.689  −0.913  .005582  My body is not understandable  29  −0.091  −2.132  0  My body is unpredictable  30c,d  −0.055  −1.794  0  My body is tense without me knowing why  31b,c  0.1  −2.266  0  I am tense in social contexts  32b  0.04  −1.278  .005275  I struggle to relax  33b,c  0.203  −2.342  0  My hands are cold and wet when I feel uncertain  34  −0.26  0.142  .816596  I relax more easily when I’m feeling safe  35f  −0.477  1.292  .000271  My muscles are often tense  36b,d  0.508  −1.886  0  I am able to sense if my muscles are tense  37  0.348  1.712  .011598  I am able to relax when I am feeling tense  38  −0.211  3.781  0  I don’t like to be touched  39b,c,g  −0.61  −0.721  .189535  I like being touched in natural situations  40  0.467  1.163  .000001  My heart rate is high even though I am not exhausted  41b  −0.229  −1.487  .001797  I feel dizzy not for any reason  42  −0.147  −1.85  .007996  I feel sick to my stomach, when faced with what I dislike  43g  −0.225  0.198  .525221  My digestion is affected by how I feel  44c,f  −0.058  4.567  0  My body always hurts  45d  −0.3  −2.683  0  I am able to sense body parts  46  −0.152  0.228  .663886  I do not understand why I have pain  47b  0.065  −1.097  .006182  My pain is affected by my mood  48f  −0.138  3.313  0  I can relieve my pain  49  −0.133  3.649  0  My body is affected by how I feel  50c,f  −0.033  1.89  .001665  My body feels different when I am happy and when I am sad  51f  −0.342  3.225  0  I try not to show how I’m feeling  52b,c  −0.212  −0.703  .039561  My body feels like an alien  53b  0.143  −1.646  0  I dislike my body  54b,g  −0.354  −2.12  0  I tense up in order to hide my feelings  55  −0.026  −2.295  0  I am ashamed of my body  56b,g  −0.458  −1.64  .000143  I am ashamed of certain parts of my body  57g  −0.212  −1.562  .003217  My feelings are expressed through my body  58  0.077  1.161  .016487  I avoid sensing my body  59b,c  −0.388  −1.607  .000019  I don’t realize that I am tired until I am exhausted  60b,e  −0.056  −1.381  .00215  It is not important to consider bodily needs  61b,e  −0.562  0.12  .479489  I know how to find new strength  62  0.137  4.717  0  I take a break when I feel tired  63  0.187  3.472  0  I always push myself to my limits  64b,e  0.066  −0.967  .000616  Body signals help me to find my limits  65  0.23  4.096  0  I do everything to manage my life on a daily basis, but still I feel bad  66b,d  0.158  −1.985  0  Description  Item No.  Location  Fit Residual  Probability  I am stable on my feet  1  0.594  3.222  0  In standing my feet have good contact with the ground  2  0.673  2.347  0  I pull myself together in standing  3  0.327  0.89  .894265  I am not aware of how I am standing  4  −0.054  1.81  .117153  I sense if my joints are flexed or extended  5  0.27  2.273  0  I can sense if my joints are tense or flexible  6  0.614  0.468  .00439  I cannot find comfortable positions in standing  7b  −0.058  −1.467  0  I am often tense  8b,c,d  0.276  −1.449  0  I am not aware of my habit positions  9  0.243  1.668  .003417  I never sit comfortably  10b,d  −0.115  −1.957  0  I can’t get comfortable when I’m lying down  11b,c  −0.197  −1.655  .000561  I search for comfortable positions  12  0.697  −0.6  .338519  My body sensation helps me find comfortable positions  13  0.854  0.663  .041528  I am always holding my breath  14b  −0.264  −1.642  .000096  I am not aware of the way I breathe  15b,c,e  0.106  0.29  .670769  I breathe more easily when I’m in a good mood  16f  −0.621  1.872  0  I try to control my breath  17b  −0.197  −0.458  .001362  I try to breathe correctly  18  0.198  1.3  .177791  I am moving smoothly  19  −0.137  4.945  0  When I am aware that I am holding breath I relax  20  0.44  0.763  .017304  My body is functioning well on a daily basis  21  −0.02  6.133  0  I always hurry even when I don’t have to  22e  0.035  0.375  .72075  I avoid being exhausted if I do not have to  23  −0.184  0.258  .6167  I am fond of moving  24  0.496  2.18  0  I don’t pay attention to the way I move  25b,c,e  −0.117  0.986  .600102  I have problems moving my arms and legs  26b  −0.381  −1.329  .001126  I move to improve my mood  27  0.214  1.867  .004535  I am afraid of moving  28b,g  −0.689  −0.913  .005582  My body is not understandable  29  −0.091  −2.132  0  My body is unpredictable  30c,d  −0.055  −1.794  0  My body is tense without me knowing why  31b,c  0.1  −2.266  0  I am tense in social contexts  32b  0.04  −1.278  .005275  I struggle to relax  33b,c  0.203  −2.342  0  My hands are cold and wet when I feel uncertain  34  −0.26  0.142  .816596  I relax more easily when I’m feeling safe  35f  −0.477  1.292  .000271  My muscles are often tense  36b,d  0.508  −1.886  0  I am able to sense if my muscles are tense  37  0.348  1.712  .011598  I am able to relax when I am feeling tense  38  −0.211  3.781  0  I don’t like to be touched  39b,c,g  −0.61  −0.721  .189535  I like being touched in natural situations  40  0.467  1.163  .000001  My heart rate is high even though I am not exhausted  41b  −0.229  −1.487  .001797  I feel dizzy not for any reason  42  −0.147  −1.85  .007996  I feel sick to my stomach, when faced with what I dislike  43g  −0.225  0.198  .525221  My digestion is affected by how I feel  44c,f  −0.058  4.567  0  My body always hurts  45d  −0.3  −2.683  0  I am able to sense body parts  46  −0.152  0.228  .663886  I do not understand why I have pain  47b  0.065  −1.097  .006182  My pain is affected by my mood  48f  −0.138  3.313  0  I can relieve my pain  49  −0.133  3.649  0  My body is affected by how I feel  50c,f  −0.033  1.89  .001665  My body feels different when I am happy and when I am sad  51f  −0.342  3.225  0  I try not to show how I’m feeling  52b,c  −0.212  −0.703  .039561  My body feels like an alien  53b  0.143  −1.646  0  I dislike my body  54b,g  −0.354  −2.12  0  I tense up in order to hide my feelings  55  −0.026  −2.295  0  I am ashamed of my body  56b,g  −0.458  −1.64  .000143  I am ashamed of certain parts of my body  57g  −0.212  −1.562  .003217  My feelings are expressed through my body  58  0.077  1.161  .016487  I avoid sensing my body  59b,c  −0.388  −1.607  .000019  I don’t realize that I am tired until I am exhausted  60b,e  −0.056  −1.381  .00215  It is not important to consider bodily needs  61b,e  −0.562  0.12  .479489  I know how to find new strength  62  0.137  4.717  0  I take a break when I feel tired  63  0.187  3.472  0  I always push myself to my limits  64b,e  0.066  −0.967  .000616  Body signals help me to find my limits  65  0.23  4.096  0  I do everything to manage my life on a daily basis, but still I feel bad  66b,d  0.158  −1.985  0  aAs presented to the first sample of participants (N = 300) and showing the disposition of the items. bItems remaining after the first Rasch analysis. cItems included in the revised BARQ. dItems included in the original BARQ subscale Function. eItems included in the original BARQ subscale Mood. fItems included in the original BARQ subscale Feelings. gItems included in the original BARQ subscale Awareness. View Large Participants A convenience sampling strategy was applied (Tab. 2). The participants were people who had musculoskeletal and/or psychosomatic complaints who consulted their primary physician for their complaints. After examination the physician referred them to NPMP treatment. Taking the inclusion and exclusion criteria into consideration, 6 NPMP specialists recruited participants consecutively to the study. Five specialists worked in primary health care, and 1 worked in an outpatient hospital clinic. The inclusion criteria were long-lasting (>3 months) musculoskeletal pain and/or psychosomatic disorders and reading skills in Norwegian. Exclusion criteria were psychiatric disorders like severe depression and anxiety. Data was collected for survey 1 in 2008, for survey 2 in 2013, and for survey 3 in 2016. Table 2. Participant Characteristics Characteristic  Survey 1 (N = 300)a  Survey 2 (n = 125)  Survey 3 (n = 48)  Sex: no. (%) of women  231 (77)  91 (73)  45 (94)  Age, y, (SD)  42.4 (11.1)  42.1 (11.0)  46.9 (12.0)  Duration of pain, y, (SD)  8.2 (7.8)  8.9 (7.7)  11.4 (7.9)  Type of work (% of participants)  Blue collar  50  43  39.6  White collar  50  57  60.4  Characteristic  Survey 1 (N = 300)a  Survey 2 (n = 125)  Survey 3 (n = 48)  Sex: no. (%) of women  231 (77)  91 (73)  45 (94)  Age, y, (SD)  42.4 (11.1)  42.1 (11.0)  46.9 (12.0)  Duration of pain, y, (SD)  8.2 (7.8)  8.9 (7.7)  11.4 (7.9)  Type of work (% of participants)  Blue collar  50  43  39.6  White collar  50  57  60.4  a14% without musculoskeletal problems. View Large Item Generation The items included in the first Rasch analysis had been collected during an extensive process with focus group interviews including NPMP physical therapists (2 groups) and people with long-lasting musculoskeletal pain (4 groups). The generated items were supplemented and commented on in written by NPMP physical therapists from different parts of Norway. All collected items were then field tested on a sample of 30 people with long-lasting musculoskeletal pain. The total item pool finally included 66 items, all refined into statements to be rated on a 7-point response scale, ranging from 1 (totally agree) to 7 (totally disagree). Twenty-four of these items, 6 in each of 4 subscales, were included in the initial BARQ (further elaborated in Dragesund et al11). First Survey The 66 items (same as above) were scored on 7-point response scales by a sample of 300 people with long-lasting musculoskeletal pain. Rasch analyses were performed to explore the possibility for obtaining a unidimensional scale. Second Survey A second survey was performed to examine whether the results from the first Rasch analysis could be confirmed in a new sample of people (n = 127). For the sake of content coverage, we included all items supported by the first Rasch analysis, plus items from the original 4 subscales of BARQ that were not supported by the first Rasch analysis, as well as a few items that were considered important from the original pool of 66 items. Third Survey Test-retest reliability was based on data from a new sample of 48 people with long-lasting musculoskeletal pain. These participants completed the revised version of the BARQ from step 2, twice within 1 week. Data Analysis The demographic and item scores were entered into SPSS (versions 17 and 21; SPSS, Chicago, Illinois) and RUMM2020 and RUMM2030 software packages (RUMM Laboratory; www.rummlab.com). The Rasch model assumes that the probability of a person affirming a trait, like in this case body awareness, in an item of a measure depends on the person's level of that trait (θ) and the level of body awareness required by the item (b). In a model in which there are several response categories, the model is expressed in a formula.16 ln (Pni/1-Pni) = θn-bi-τi. Pni is the probability that a person n will affirm the item, θn is the person's level of the trait, bi is the level of the trait expressed by the item, and τi represents the .5 probability point (threshold) between adjacent response levels for that item. In this case it is assumed that as a person's body awareness increases, the probability of a maximum score on the item decreases. Person and item scores are used to “calibrate” items on a logit scale. Items at one end of the scale are “easier,” whereas items at the other end are “more difficult.” In the present analyses, items with a positive calibration were more difficult (ie, indicated less body awareness). A logit is the natural log-odds of the level of difficulty of a particular item in relation to all other items in the scale (hierarchy). Rasch analysis also constructs a hierarchy of the respondents; the hierarchy was ordered by the respondents’ level of disability. Unidimensionality refers to the single underlying construct measured by items that form a scale; for this instrument body awareness. The adequacy of the fit of each item to the Rasch model is assessed by the overall model fit, individual person fit and item fit, thresholds of the response categories, local dependency, differential item functioning (DIF), and unidimensionality. This requires a number of tests for effecting a transformation of the ordinal, raw scores into linear, interval level measures.17,18 The overall fit of data to the model was examined with item-trait interaction statistics. A nonsignificant chi-square probability value, Bonferroni adjusted, indicates the hierarchical ordering of the items to be consistent across all levels of the trait, body awareness. The fit of individual items and persons was examined by inspecting their fit residuals, the difference between their observed and expected logit values. If the residuals were beyond the standardized ±2.5 (99% CI) or if an item showed a significant chi-square probability value, misfit was considered. The mean of all fit residual across all items of the scale and across all person estimates should be closed to 0 with a SD close to 1 (preferably < 1.40). The local response independency was considered to be dependent if the residual correlation was greater than .2 above the average of all residual correlations.19 Unidimensionality The concept of unidimensionality is a primary tenet underpinning the Rasch analysis. If the scores on the revised version of the BARQ are to be added into a total score representing an overall level of body awareness, the analysis should indicate a unidimensional scale, measuring 1 construct. The unidimensionality was tested by contrasting items loading positively and negatively on the first principle component of the residuals, through series of independent t tests.20 The present of significant multidimensionality was noted if more than 5% of these t tests were significant or if the lower CI of the observed proportion was below 0.05.21 Response threshold ordering In Rasch analysis the response options for each item should reflect a logical progression of the underlying trait, body awareness. If this is not demonstrated, thresholds are disordered and can be re-scored by collapsing categories.15 DIF To explore potential bias across groups of respondents, analysis for DIF was performed for 3 factors: sex, age (2 groups: ≥ 43 years and < 43 years), and duration of pain (2 groups: ≥ 5 years and <5 years). Rasch analysis generates a Person Separation Index, which is a reliability value equivalent to the Cronbach alpha for internal consistency. A Person Separation Index of greater than 0.70 was considered desirable for comparing groups and also as evidence of sufficient discriminative ability.22 Targeting of Persons and Items In clinical practice it is important that the items are appropriately targeted for the population for whom the questionnaire is developed. Poorly targeted questionnaires result in floor and ceiling effects. Scale-to-sample targeting was made by comparing person locations from the sample with locations of the scale, requiring means close to 0 and an SD close to 1.15 This might also be visualized by a person-item threshold distribution histogram, which informs about the suitability of the sample for evaluating the scale and the suitability of the scale for measuring the sample. Test-Retest Reliability The COnsensus-based Standards for the selection of health status Measurement InstrumeNts (COSMIN) guidelines were followed.23 The intraclass correlation coefficient [ICC(2.1)] was used to examine test-retest reliability.24 An ICC of .80 or more was considered evidence of satisfactory test-retest reliability. Repeatability was assessed by calculating the standard error of measurement between scores at test and retest, being the square root of the mean square error (total in the analysis-of-variance table). The 95% CI for repeatability was further calculated by multiplying the standard error of measurement by 2.77.25 This value is often referred to as the minimal detectable change, being the smallest change that must be observed between 2 measurements before it can be considered to exceed measurement error with 95% certainty. Role of the Funding Source The Norwegian Fund for Postgraduate Training in Physiotherapy provided financial support of the study. The funding source had no role in the study. Results Characteristics of Participants Characteristics of the participants taking part in the 3 surveys are presented in Table 2. The first survey included 300 participants, mostly women (n = 231). The mean age was 42.4 years. Most participants (86%) were treated for long-lasting musculoskeletal and psychosomatic disorders, with an average duration of pain of 8.2 years. The remaining participants (14%), being healthy without musculoskeletal pain problems. In the second survey there were 125 participants, also mostly women (n = 91). The mean age was 42.1 years. They all had long-lasting musculoskeletal pain, with average duration of pain of 8.9 years. The third survey included 48 (93.8% women), all with long-lasting musculoskeletal pain. The mean age was 46.9 years, and their mean duration of pain was 11.4 years. Survey 1: First Rasch Analysis of Initial Item Pool All 66 items were scored on a 7-point scale by 300 people in the first survey (Tab. 1). The missing data on the initial 66 items ranged from 0.3% to 11%. For Rasch analysis of polytomous scales there are 2 models to be used: the unrestricted model or the rating scale model. The unrestricted model places no constraints on the threshold parameters, whereas the rating scale model constrains all thresholds to be equally spaced across the trait of the items. Results on the likelihood ratio test decide which model to choose. The likelihood ratio test was significant (P < .0000001), suggesting that the unrestricted model should be adopted. The overall statistics indicated that the 66 items did not reflect a unidimensional measure (P = .00076). The analysis showed misfit of both persons and items outside the fit residual range of ±2.5. This might have been due to the fact that all the 66 items had disordered thresholds. All items had to be rescored to a 3-point response scale in order to avoid disordered thresholds (scores of 1 and 2 = 0, scores of 3 to 5 = 1, and scores of 6 and 7 = 2). Only 4 of the items (40, 42, 43, and 57) had disordered thresholds after the rescoring, suggesting that the 3-point scale worked much better than the 7-point scale. The overall statistics were still highly significant (P = .000). Fifty-two participants had extreme scores outside the ±2.5 limit, 1 item (55) was redundant (fit residual = −2.63), and 10 items (19, 21, 38, 44, 48, 49, 51, 62, 63, and 65) had fit residuals above 2.5. In addition, several of the items had poor statistics in terms of a significant chi-square probability value (adjusted for Bonferroni: .00076) (items 29, 54, 33, 45, 31, 10, 36, 7, 8, 40, 16, 2, 5, 9, 24, and 1). Removal of misfitting and redundant items were carried out during 11 following analyses. The process ended up with a unidimensional 26-item version (Tab. 3). Thirteen of these 26 items overlapped with items from the original BARQ, while eleven were excluded during the analysis. When checking the DIF of this version, a DIF for age group was found for items 8 and 56, and a DIF for duration of pain was found for items 15, 36, and 66. There was no DIF for sex. Table 3. Twenty-six Items Resulting From First Rasch Analysisa Item Description  Item No.b  Item Difficulty (Logits)  Standard Error for Location  Fit Residual  Probability  My muscles are often tense  36  2.269  0.508  −1.886  0  I am often tense  8  1.273  0.276  −1.449  0  I do everything to manage my life on a daily basis, but still I feel bad  66  1.137  0.158  −1.985  0  I struggle to relax  33  0.955  0.203  −2.342  0  I try not to show my feelings  52  0.763  0.143  −0.703  .039561  My body feels tense and I do not know why  31  0.753  0.1  −2.266  0  I always push myself to my limits  64  0.606  0.066  4.096  0  I am not aware of the way I breathe  15  0.489  0.106  0.29  .670769  I am tense in social contexts  32  0.435  0.04  −1.278  .005275  I try to control my breath  17  0.142  −0.197  −0.458  .001362  I cannot find comfortable positions in standing  7  0.13  −0.058  −1.467  0  I don’t realize that I am tired until I am exhausted  60  0.086  −0.056  −1.381  .002150  I never sit comfortably  10  0.026  −0.115  −1.957  0  I can’t get comfortable when I’m lying down  11  −0.077  −0.197  −1.655  .000561  I do not understand why I have pain  47  −0.168  −0.138  −1.097  .006182  I am always holding my breath  14  −0.212  −0.264  −1.642  .000096  I don’t pay attention to the way I move  25  −0.217  −0.117  0.986  .600102  My heart rate is high even though I am not exhausted  41  −0.379  −0.229  −1.487  .001797  I dislike my body  54  −0.474  −0.3  −2.12  0  I avoid sensing my body  59  −0.567  −0.388  −1.607  .000019  I have problems moving my arms and legs  26  −0.69  −0.381  −1.329  .001126  My body feels like an alien  53  −0.771  −0.354  −1.646  0  I am ashamed of my body  56  −1.071  −0.458  −1.64  .000143  It is not important to consider bodily needs  61  −1.357  −0.562  0.12  .479489  I don’t like to be touched  39  −1.467  −0.61  −0.721  .189535  I am afraid of moving  28  −1.615  −0.689  −0.913  .005582  Item Description  Item No.b  Item Difficulty (Logits)  Standard Error for Location  Fit Residual  Probability  My muscles are often tense  36  2.269  0.508  −1.886  0  I am often tense  8  1.273  0.276  −1.449  0  I do everything to manage my life on a daily basis, but still I feel bad  66  1.137  0.158  −1.985  0  I struggle to relax  33  0.955  0.203  −2.342  0  I try not to show my feelings  52  0.763  0.143  −0.703  .039561  My body feels tense and I do not know why  31  0.753  0.1  −2.266  0  I always push myself to my limits  64  0.606  0.066  4.096  0  I am not aware of the way I breathe  15  0.489  0.106  0.29  .670769  I am tense in social contexts  32  0.435  0.04  −1.278  .005275  I try to control my breath  17  0.142  −0.197  −0.458  .001362  I cannot find comfortable positions in standing  7  0.13  −0.058  −1.467  0  I don’t realize that I am tired until I am exhausted  60  0.086  −0.056  −1.381  .002150  I never sit comfortably  10  0.026  −0.115  −1.957  0  I can’t get comfortable when I’m lying down  11  −0.077  −0.197  −1.655  .000561  I do not understand why I have pain  47  −0.168  −0.138  −1.097  .006182  I am always holding my breath  14  −0.212  −0.264  −1.642  .000096  I don’t pay attention to the way I move  25  −0.217  −0.117  0.986  .600102  My heart rate is high even though I am not exhausted  41  −0.379  −0.229  −1.487  .001797  I dislike my body  54  −0.474  −0.3  −2.12  0  I avoid sensing my body  59  −0.567  −0.388  −1.607  .000019  I have problems moving my arms and legs  26  −0.69  −0.381  −1.329  .001126  My body feels like an alien  53  −0.771  −0.354  −1.646  0  I am ashamed of my body  56  −1.071  −0.458  −1.64  .000143  It is not important to consider bodily needs  61  −1.357  −0.562  0.12  .479489  I don’t like to be touched  39  −1.467  −0.61  −0.721  .189535  I am afraid of moving  28  −1.615  −0.689  −0.913  .005582  aThe items are presented in location order, with easy items on the top and difficult items on the bottom. bFrom the initial pool of 66 items. View Large Survey 2: Second Rasch Analysis The second survey included 39 items that were scored on a 4-point response scale, from 0 (totally agree) to 3 (totally disagree); 26 items resulting from the first Rasch analysis, 11 items from the original BARQ subscales that were excluded in the first Rasch analysis, and 2 items from the original items pool that seemed clinically important. A 4-point response scale was chosen in order to improve the reliability of the scale. The missing data on these 39 items were low and ranged from 0.3% to 3.1%. The unrestricted model was initially adopted (P < .0000001) on all 127 people. The overall statistics were significant (P = .036), indicating that the initial 39 items did not reflect a unidimensional measure. The residual mean value for persons (−0.03; SD = 1.51), indicated some misfit among respondents in the sample, and individual item fit, adjusted for Bonferroni, also indicated misfit of items 40 and 62. Two participants were excluded from the analysis because they had not reported sex or age categories. Scores of the remaining 125 participants were thus included in the further analysis. Item 40 and item 62 were deleted because of fit residual of greater than 2.5. Items with local independency (items 32, 36, and 47), items with similar locations (reflecting degree of difficulty) (items 7, 14, 16, 17, 22, 26, 28, 35, 41, 43, 45, 48, 51, 53, 54, 56, 57, 60, 61, 64, and 66), as well as 1 respondent not responding as expected (id no. 113) were systematically deleted in order to improve the data fit to the Rasch model and to achieve a questionnaire with fewer items (Tab. 4). The process ended up with a 12-item solution with a 4-point response scale (Tab. 5). The Norwegian and English versions of the final revised version of the BARQ (BARQ-R) are presented in eAppendixes 1 and 2. Table 4. Thirty-nine Items Scored by 127 Participants and Included in Second Rasch Analysis Item Description  Item No.a  Item Difficulty (Logits)  Standard Error for Location  Fit Residual  Probability  Reason for Exclusion  I cannot find comfortable positions in standing  7  −0.254  0.108  1.365  .398352  Location similar to other  I am often tense  8  −1.314  0.132  0.572  .990277    I never sit comfortably  10  −0.215  0.103  3.480  .001004  Misfit > 2.5  I can’t get comfortable when I’m lying down  11  0.339  0.105  2.010  .770461    I am always holding my breath  14  0.281  0.111  0.029  .695983  Location similar to other  I am not aware of the way I breathe  15  0.433  0.124  0.325  .674593    I breathe more easily when I’m in a good mood  16  −1.123  0.110  0.596  .617333  Location similar to other  I try to control my breath  17  −0.052  0.114  1.419  .096183  Location similar to other  I always hurry even when I don’t have to  22  −0.463  0.098  −0.623  .720634  Location similar to other  I don’t pay attention to the way I move  25  0.307  0.115  0.535  .399475    I have problems moving my arms and legs  26  0.967  0.121  −0.648  .157807  Location similar to other  I am afraid of moving  28  0.834  0.117  0.738  .213278  Location similar to other  My body is unpredictable  30  0.571  0.108  0.282  .284172    My body is tense without me knowing why  31  −0.133  0.107  −0.275  .143670    I am tense in social contexts  32  0.180  0.104  −2.239  .006189  Local dependency  I struggle to relax  33  −0.305  0.109  −0.567  .774623    I relax more easily when I’m feeling safe  35  0.780  0.099  1.045  .393768  Location similar to other  My muscles are often tense  36  −1.020  0.115  0.264  .765420  Local dependency  I don’t like to be touched  39  1.314  0.129  0.340  .838816    I like being touched in natural situations  40  0.490  0.100  2.923  .000072  Misfit > 2.5  My heart rate is high even though I am not exhausted  41  0.683  0.109  1.189  .032264  Location similar to other  I feel sick to my stomach, when faced with what I dislike  43  0.746  0.111  0.352  .746936  Location similar to other  My digestion is affected by how I feel  44  0.050  0.097  0.357  .545231    My body always hurts  45  −0.082  0.109  −0.651  .370472  Location similar to other  I do not understand why I have pain  47  0.567  0.116  −0.394  .542590  Local dependency  My pain is affected by my mood  48  −0.661  0.106  1.431  .552783  Location similar to other  My body is affected by how I feel  50  −0.710  0.111  1.121  .417151    My body feels different when I am happy and when I am sad  51  −0.524  0.102  −0.007  .339661  Location similar to other  I try not to show how I’m feeling  52  −0.003  0.100  −0.123  .965726    My body feels like an alien  53  0.944  0.115  −0.029  .493674  Location similar to other  I dislike my body  54  0.458  0.109  −0.961  .187169  Location similar to other  I am ashamed of my body  56  1.002  0.116  −0.225  .233640  Location similar to other  I am ashamed of certain parts of my body  57  0.328  0.099  0.210  .328811  Location similar to other  I avoid sensing my body  59  0.931  0.125  −0.017  .915166    I don’t realize that I am tired until I am exhausted  60  0.117  0.105  0.594  .104907  Location similar to other  It is not important to consider bodily needs  61  1.012  0.118  0.588  .224231  Location similar to other  I always push myself to my limits  64  0.277  0.104  1.703  .259346  Location similar to other  I do everything to manage my life on a daily basis, but still I feel bad  66  −0599  0.108  0.539  .174838  Location similar to other  I know how to find new strength  62  0.373  0.117  2.553  .001205  Misfit > 2.5  Item Description  Item No.a  Item Difficulty (Logits)  Standard Error for Location  Fit Residual  Probability  Reason for Exclusion  I cannot find comfortable positions in standing  7  −0.254  0.108  1.365  .398352  Location similar to other  I am often tense  8  −1.314  0.132  0.572  .990277    I never sit comfortably  10  −0.215  0.103  3.480  .001004  Misfit > 2.5  I can’t get comfortable when I’m lying down  11  0.339  0.105  2.010  .770461    I am always holding my breath  14  0.281  0.111  0.029  .695983  Location similar to other  I am not aware of the way I breathe  15  0.433  0.124  0.325  .674593    I breathe more easily when I’m in a good mood  16  −1.123  0.110  0.596  .617333  Location similar to other  I try to control my breath  17  −0.052  0.114  1.419  .096183  Location similar to other  I always hurry even when I don’t have to  22  −0.463  0.098  −0.623  .720634  Location similar to other  I don’t pay attention to the way I move  25  0.307  0.115  0.535  .399475    I have problems moving my arms and legs  26  0.967  0.121  −0.648  .157807  Location similar to other  I am afraid of moving  28  0.834  0.117  0.738  .213278  Location similar to other  My body is unpredictable  30  0.571  0.108  0.282  .284172    My body is tense without me knowing why  31  −0.133  0.107  −0.275  .143670    I am tense in social contexts  32  0.180  0.104  −2.239  .006189  Local dependency  I struggle to relax  33  −0.305  0.109  −0.567  .774623    I relax more easily when I’m feeling safe  35  0.780  0.099  1.045  .393768  Location similar to other  My muscles are often tense  36  −1.020  0.115  0.264  .765420  Local dependency  I don’t like to be touched  39  1.314  0.129  0.340  .838816    I like being touched in natural situations  40  0.490  0.100  2.923  .000072  Misfit > 2.5  My heart rate is high even though I am not exhausted  41  0.683  0.109  1.189  .032264  Location similar to other  I feel sick to my stomach, when faced with what I dislike  43  0.746  0.111  0.352  .746936  Location similar to other  My digestion is affected by how I feel  44  0.050  0.097  0.357  .545231    My body always hurts  45  −0.082  0.109  −0.651  .370472  Location similar to other  I do not understand why I have pain  47  0.567  0.116  −0.394  .542590  Local dependency  My pain is affected by my mood  48  −0.661  0.106  1.431  .552783  Location similar to other  My body is affected by how I feel  50  −0.710  0.111  1.121  .417151    My body feels different when I am happy and when I am sad  51  −0.524  0.102  −0.007  .339661  Location similar to other  I try not to show how I’m feeling  52  −0.003  0.100  −0.123  .965726    My body feels like an alien  53  0.944  0.115  −0.029  .493674  Location similar to other  I dislike my body  54  0.458  0.109  −0.961  .187169  Location similar to other  I am ashamed of my body  56  1.002  0.116  −0.225  .233640  Location similar to other  I am ashamed of certain parts of my body  57  0.328  0.099  0.210  .328811  Location similar to other  I avoid sensing my body  59  0.931  0.125  −0.017  .915166    I don’t realize that I am tired until I am exhausted  60  0.117  0.105  0.594  .104907  Location similar to other  It is not important to consider bodily needs  61  1.012  0.118  0.588  .224231  Location similar to other  I always push myself to my limits  64  0.277  0.104  1.703  .259346  Location similar to other  I do everything to manage my life on a daily basis, but still I feel bad  66  −0599  0.108  0.539  .174838  Location similar to other  I know how to find new strength  62  0.373  0.117  2.553  .001205  Misfit > 2.5  aFrom the initial pool of 66 items. View Large Table 5. Final Revised Body Awareness Rating Questionnaire, With 4-Point Response Scalea Item Description  Completely 
Disagree (0)  Somewhat 
Disagree (1)  Somewhat 
Agree (2)  Completely 
Agree (3)  1. I am often tenseb          2. My body is affected by how I feelb          3. I am not aware of the way I breatheb          4. I don’t pay attention to the way I moveb          5. I struggle to relax          6. My body is tense without me knowing why          7. I try not to show how I’m feeling          8. My digestion is affected by how I feelb          9. I can’t get comfortable when I’m lying down          10. My body is unpredictableb          11. I avoid paying too much attention to my body          12. I don’t like to be touchedb          Item Description  Completely 
Disagree (0)  Somewhat 
Disagree (1)  Somewhat 
Agree (2)  Completely 
Agree (3)  1. I am often tenseb          2. My body is affected by how I feelb          3. I am not aware of the way I breatheb          4. I don’t pay attention to the way I moveb          5. I struggle to relax          6. My body is tense without me knowing why          7. I try not to show how I’m feeling          8. My digestion is affected by how I feelb          9. I can’t get comfortable when I’m lying down          10. My body is unpredictableb          11. I avoid paying too much attention to my body          12. I don’t like to be touchedb          aThe items are presented in location order, with easier items on the top and more difficult items on the bottom. bAlso included in the original Body Awareness Rating Questionnaire. View Large The final 12-item solution showed a good fit to the Rasch model, with a chi-square value of 13.63 and a P value of .954 (Bonferroni adjusted 0.5/12 = .004). All the items had satisfactory fit residual, in the range of −2.010 to 0.567 and the item difficulty levels ranged from 1.314 (most difficult items) to −1.314 (least difficult items). The Person Separation Index was 0.76. None of the 12 items had disordered threshold. Furthermore, no significant DIF was demonstrated by sex, age groups and duration of pain, making further adjustment unnecessary. And finally the independent t test between the 2 subsets of items with positive and negative loadings on the first residual component was not statistically significant (P = .159), supporting the unidimensionality of the 12 items. Among these 12 items, 7 items were included in the 4 subscales of the original BARQ (items 8, 50, 30, 44, 39, 15, 25, and 2). Additionally, 9 of the 12 items (8, 15, 25, 33, 31, 52, 11, 59, and 39) were confirmed from the result of the first Rasch analysis (Tab. 1). The 12 items also seem to target the population for whom they are developed (mean person location = −0.04; SD = 0.74) (eFigure; available at https://academic.oup.com/ptj). Survey 3: Test-Retest Reliability The mean scores on test and retest were 20.6 (SD = 5.5) and 20.2 (SD = 5.4), respectively. The mean difference between test and retest for the 12-item scale (sum score ranging from 0 to 36) was 0.44 (SD = 3.2), and the ICC was .83 (95% CI = .71–.89). The standard error of measurement was 2.26, and the smallest detectable change with 95% CI was 6.26, or 17.4% of the total score of 36. This implies that a change in individual participants should be above this value to indicate a change above measurement error. Discussion In the present study, we succeeded in developing a unidimensional assessment tool of body awareness based on previously collected test items for BARQ by using Rasch analysis. The BARQ-R includes 12 self-reported items scored on a 4-point ordinal scale (0–3), reflecting degrees of body awareness even though the items included in this solution is slightly different from the 4-factor solution (original BARQ). The scale also seems to express different levels of problems with body awareness as seen in the clinic, from the most common problems (feeling tense and be affected by mood) to the most challenging (feeling that the body is unpredictable and not liking to be touched). The results from the study indicate that the BARQ-R is a valid and reliable assessment tool that can be used to capture body awareness as reported by people with musculoskeletal pain. For comparison the original BARQ included 24 items in 4 subscales (6 items in each subscale) that could not be combined into a sum score. The BARQ-R is briefer and more user-friendly, taking only 3 to 5 minutes to complete. We therefore recommend applying the BARQ-R for further research and clinical use. The BARQ-R includes a variety of items concerning different aspects of body awareness, like awareness of muscle tension, breath, movements, and the experience of coherence between body and emotions.2 Many physical therapists specialized in a body-mind approach of physical therapy (NPMP) contributed to the generation of items, as well as a large target group of people with musculoskeletal and/or psychosomatic problems. Physical therapists and patients might perceive the phenomenon of body awareness differently, and commonly clinical measurements are based on clinician's opinion only and therefore lack content validity from the perspective of patients. Thus, a major strength of the BARQ-R as well as of the original BARQ is content validity with items derived from both the physical therapists’ and the patients’ perspectives. This is in line with quality criteria recommended by the COSMIN group, underscoring the need to include both patients and therapists when constructing a questionnaire.26 The demographic characteristics of the sample were accordingly quite similar to those of patients on waiting-lists for NPMP described by Breitve et al.27 In the present study Rasch analysis was used to develop a unidimensional measure, which is considered an essential quality when summing up individual items to obtain a valid total score.13,14 The Rasch model enables analysis of targeting of items’ difficulty to persons’ ability by calibrating items and persons on a common scale with interval level units (logits). As such, a primary strength of Rasch analysis is that ordinal scores may be converted into interval level scores. Furthermore the method makes it possible to analyze whether items exhibit a consistent and invariant hierarchy of difficulty,13,14 and if response options for each item reflect a logical progression of the underlying trait.15 The first Rasch analyses clearly showed that the original 7-point response scale did not function for any of the 66 items. Threshold disordering was avoided by collapsing them into a 3-point response scale. This aspect of scoring alternatives of the response scale was not examined in the original BARQ, which was developed based on classical test theory. A scale with disordered threshold will be problematic when used to evaluate effect of an intervention, because it will be difficult to evaluate a change in categories.28 Since it was obvious that the participants were not able to distinguish between 7 levels of body awareness (7-point response scale), we used a 3-point scale in the first Rasch analysis by collapsing scoring alternatives, and found it to be appropriate. However, as we considered a 4-point response scale to be more informative and sensitive to differences in the trait than a 3-point scale, we tried out a 4-point scale when new data were collected in the second Rasch analysis, showing satisfactory results. This is in line with Streiner and Norman,22 who stated that an increase in the number of response options on a rating scale will increase the information value of any observation and improve the reliability, although the agreement will decrease. The 4-point response scale in the second Rasch analysis was shown to improve the reliability of the scale as compared to the results of the first Rasch analysis with a 3-point response scale. Nine of 26 items from the first Rasch analysis (based on 66 items) were confirmed in the second Rasch analysis (based on 39 items), and 7 items of the final 12-item scale (BARQ-R), represented all 4 subscales of the original BARQ (2 from the subscale Function, 2 from the subscale Mood, 1 from the subscale Feelings, and 2 from the subscale Awareness) (Tabs. 1 and 5). The items are also ranked hierarchically according to difficulties of the phenomenon of body awareness across age groups, sexes and duration of pain. The evidence of test-retest reliability was also confirmed, giving an ICC of greater than .80. The smallest detectable change, however, was high, being 6.26, or 17.4% of the total score. Still, the BARQ-R has undergone an extensive improvement in measurement properties. Limitations of the current study concern the sample size in the second survey. Linacre29 has, when using Rasch analysis, proposed a sample size of 250 in order to achieve appropriate degree of precision. The first Rasch analysis included a group of 300 target people, in line with these recommendations. In the second Rasch analysis, however, the sample size was only 125, below the recommended number. To conclude, the present study provides support for a satisfactory internal construct validity and reliability of the 12-item BARQ-R. Thus, the BARQ-R can be recommended used by physical therapists focusing on body awareness among people with long-lasting musculoskeletal pain. Further research to assess other aspects of validity and responsiveness should be carried out. Author Contributions and Acknowledgments Concept/idea/research design: L.I. Strand, M. Grotle Writing: T. Dragesund, L.I. Strand, M. Grotle Data collection: T. Dragesund Data analysis: T. Dragesund, M. Grotle Project management: T. Dragesund Providing facilities/equipment: L.I. Strand Providing institutional liaisons: L.I. Strand Consultation (including review of manuscript before submitting): L.I. Strand, M. Grotle The authors thank the Norwegian Fund for Postgraduate Training in Physiotherapy for financial support of the study. They also thank the physical therapists who recruited participants for the study as well as the participants. Ethics Approval The study was approved by the regional ethics committee (REK West 2012/2238). References 1 Kvåle A, Ljunggren AE. Body awareness therapies. In: Schmidt RF, Willis WD, eds. Encyclopedia of Pain . Berlin, Germany: Springer Verlag; 2007: 167– 169. 2 Thornquist E, Bunkan BH. What Is Psychomotor Physiotherapy?  Oslo, Norway: Norwegian University Press; 1991. 3 Øien AM, Råheim M, Iversen S, Steihaug S. Self perception as embodied knowledge: changing processes for patients with chronic pain. Adv Physiother . 2009; 11: 121– 129. Google Scholar CrossRef Search ADS   4 Dragesund T, Raheim M. Norwegian psychomotor physiotherapy and patients with chronic pain: patients’ perspective on body awareness. Physiother Theory Pract . 2008; 24: 243– 254. Google Scholar CrossRef Search ADS PubMed  5 Roxendal G. Body Awareness Therapy and the Body Awareness Scale: Treatment and Evaluation in Physiotherapy [dissertation] . Gothenborg, Sweden: University of Gothenborg; 1985. 6 Skjærven LH, Kristoffersen K, Gard G. How can movement quality be promoted in clinical practice? A phenomenological study of physical therapist experts. Phys Ther . 2010; 90: 1– 15. Google Scholar CrossRef Search ADS   7 Bruchon-Schweitzer M. Dimensionality of the body-image: the Body-Image Questionnaire. Percept Mot Skills . 1987; 65: 887– 892. Google Scholar CrossRef Search ADS   8 Slade PD, Dewey ME, Newton T, Brodie D, Kiemle G. Development and preliminary validation of the Body Satisfaction Scale (BSS). Psychol Health . 1990; 4: 213– 220. Google Scholar CrossRef Search ADS   9 Rosen JC, Srebnik D, Saltzberg E, Wendt S. Development of a body image avoidance questionnaire. Psychol Assess . 1991; 3: 32– 37. Google Scholar CrossRef Search ADS   10 Cash TF, Szymanski ML. The development and validation of the Body-Image Ideals Questionnaire. J Pers Assess . 1995; 64: 466– 477. Google Scholar CrossRef Search ADS PubMed  11 Dragesund T, Ljunggren AE, Kvåle A, Strand LI. Body Awareness Rating Questionnaire: development of a self-administered questionnaire for patients with long-lasting musculoskeletal and psychosomatic disorders. Adv Physiother . 2010; 12: 87– 94. Google Scholar CrossRef Search ADS   12 Dragesund T, Målfrid R, Strand LI. Body Awareness Rating Questionnaire: measurement properties. Physiother Theory Pract . 2012; 28: 515– 528. Google Scholar CrossRef Search ADS PubMed  13 Rasch G. Probabilistic Models for Some Intelligence and Attainment Tests . Chicago, IL: MESA Press; 1980. 14 Karabatsos G. The Rasch model, additive conjoint measurement, and new models of probabilistic measurement theory. J Appl Meas . 2001; 2: 389– 423. Google Scholar PubMed  15 Tennant A, Conaghan PG. The Rasch measurement model in rheumatology: what is it and why use it? When should it be applied, and what should one look for in a Rasch paper? Arthritis Rheum . 2007; 57: 1358– 1362. Google Scholar CrossRef Search ADS PubMed  16 Tennant A, McKenna SP, Hagell P. Application of Rasch analysis in the development and application of quality of life instruments. Value Health . 2004; 7: 22– 26. Google Scholar CrossRef Search ADS PubMed  17 Pallant JF, Tennant A. An introduction to the Rasch measurement model: an example using the Hospital Anxiety and Depression Scale (HADS). Br J Clin Psychol . 2007; 46: 1– 18. Google Scholar CrossRef Search ADS PubMed  18 Grimby G, Tennant A, Tesio L. The use of raw scores from ordinal scales: time to end malpractice? J Rehabil Med . 2012; 44: 97– 98. Google Scholar CrossRef Search ADS PubMed  19 Marais I, Andrich D. Formalizing dimension and response violations of local independence in the unidimensional Rasch model. J Appl Meas . 2008; 9: 200– 215. Google Scholar PubMed  20 Smith EVJr. Detecting and evaluating the impact of multidimensionality using item fit statistics and principal component analysis of residuals. J Appl Meas . 2002; 3: 205– 231. Google Scholar PubMed  21 Tennant A, Pallant JF. Unidimensionality Matters! (A Tale of Two Smiths?). Rasch Meas . 2006; 20: 1048– 1051. 22 Streiner DL, Norman GR. Health Measurement Scales: A Practical Guide to Their Development and Use . Oxford, England: Oxford Medical Publications; 2008. Google Scholar CrossRef Search ADS   23 Mokkink LB, Terwee CB, Knol DL et al.   The COSMIN checklist for evaluating the methodological quality of studies on measurement properties: a clarification of its content. BMC Med Res Methodol . 2010; 10: 22. Google Scholar CrossRef Search ADS PubMed  24 Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull . 1979; 86: 420– 428. Google Scholar CrossRef Search ADS PubMed  25 Bland JM, Altman DG. Measurement error. BMJ . 1996; 312: 1654. Google Scholar CrossRef Search ADS PubMed  26 Mokkink LB, Terwee CB, Patrick DL et al.   The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes. J Clin Epidemiol . 2010; 63: 737– 745. Google Scholar CrossRef Search ADS PubMed  27 Breitve MH, Hynninen M, Kvåle A. Emosjonelle symptomer og helseplager hos pasienter som oppsøker psykomotorisk fysioterapi [Emotional symptoms and health problems in patients consulting Norwegian psychomotor physiotherapy]. Fysioterapeuten . 2008; 75: 19– 26. 28 Chen WH, Lenderking W, Jin Y, Wyrwich KW, Gelhorn H, Revicki DA. Is Rasch model analysis applicable in small sample size pilot studies for assessing item characteristics? An example using PROMIS pain behavior item bank data. Qual Life Res . 2014; 23: 485– 493. Google Scholar CrossRef Search ADS PubMed  29 Linacre JM. Optimizing rating scale category effectiveness. J Appl Meas . 2002; 3: 85– 106. Google Scholar PubMed  © 2017 American Physical Therapy Association

Journal

Physical TherapyOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off