Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Establishing Linkages Between Distributed Survey Responses and Consumer Wearable Device Datasets: A Pilot Protocol

Establishing Linkages Between Distributed Survey Responses and Consumer Wearable Device Datasets:... Background: As technology increasingly becomes an integral part of everyday life, many individuals are choosing to use wearable technology such as activity trackers to monitor their daily physical activity and other health-related goals. Researchers would benefit from learning more about the health of these individuals remotely, without meeting face-to-face with participants and avoiding the high cost of providing consumer wearables to participants for the study duration. Objective: The present study seeks to develop the methods to collect data remotely and establish a linkage between self-reported survey responses and consumer wearable device biometric data, ultimately producing a de-identified and linked dataset. Establishing an effective protocol will allow for future studies of large-scale deployment and participant management. Methods: A total of 30 participants who use a Fitbit will be recruited on Mechanical Turk Prime and asked to complete a short online self-administered questionnaire. They will also be asked to connect their personal Fitbit activity tracker to an online third-party software system, called Fitabase, which will allow access to 1 month’s retrospective data and 1 month’s prospective data, both from the date of consent. Results: The protocol will be used to create and refine methods to establish linkages between remotely sourced and de-identified survey responses on health status and consumer wearable device data. Conclusions: The refinement of the protocol will inform collection and linkage of similar datasets at scale, enabling the integration of consumer wearable device data collection in cross-sectional and prospective cohort studies. (JMIR Res Protoc 2017;6(4):e66) doi: 10.2196/resprot.6513 KEYWORDS Fitbit; Mturk; mHealth; clinical research protocol; consumer wearable; physical activity tracker quality and duration of sleep, heart rate, and location [1]. Introduction Researchers would benefit from being able to remotely gather and link these data without the need for face-to-face interaction, The increasing variety, functionality, and storage capacity of encouraging the use of the respondent’s own devices in the data consumer wearable devices has created an opportunity for using collection process. the data collected on these devices for research purposes. Consumer wearables are most commonly worn on the wrist but A remote data collection protocol would reduce the cost of may also be worn on clothing, at the waist, or as part of eyewear providing devices to participants, reduce the time spent meeting or earwear. In this paper, we focus on activity trackers, a type and training respondents on their use, and increase the speed at of consumer wearable which can collect a variety of data which data could be collected [2]. Furthermore, with the ability including steps, distance, physical activity, calories burned, to efficiently collect health data remotely, hospitals, worksites, http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR RESEARCH PROTOCOLS Brinton et al and other health care providers could monitor participants in in exchange for money deposited into their Amazon.com real time, reducing the need for frequent check-ups and the personal account. With the help of Mturk Prime, this study will financial strain that is associated with those appointments [3]. use Mturkers as a basis for the sample. Mturk Prime operates The utility and success of remote-access interventions was as a team of online support staff who assist researchers in demonstrated to be effective in collecting biometric (eg, managing, communicating with, and collecting data from continuous heart rate, sleep, and other health indicators) [4] Mturkers (www.turkprime.com). Research using Mturkers has data, yet research using consumer wearable devices is limited increased dramatically in the past few years due to the low cost and presents challenges. and vast acquisition of data [9]. Additionally, studies show that Mturkers are more demographically diverse than standard While research investigating the viability and functionality of convenience samples and samples from other online forms of consumer wearables has increased in recent years, thanks in data collection such as Twitter [10] and may be generalizable part to calls for research from the National Institutes of Health to the greater population [11,12]. In order for an Mturker to and United Nations International Children's Fund, challenges become a part of this study’s panel, the person will be required still exist in the utility of these devices as the most efficient and to either keep track of their own weight, diet, or exercise routine useful way to learn about the health of an individual. Frequently, or keep track of their own blood pressure, blood sugar, sleep remote data collection may be hindered by burdensome patterns, headaches, or some other health-related indicator. An notification systems, forcing individuals to use study-provided eligible Mturker must also be at least 18 years of age, regularly devices they are not comfortable with and requiring frequent wear a Fitbit, and be willing to give the research team access face-to-face contact with participants in order to download data to their Fitbit data for the previous month and the upcoming from their devices. For example, many present studies employ month from the date of sign-up. experience sampling methodology, which requires that researchers notify participants throughout the day requesting Study Design and Procedure that they provide their remote data, a burdensome process for The sample will consist of 30 Mturk participants. Participants both respondents and researchers [5,6,7]. They often frequently will read a short task description and compensation information require that participants travel to the researcher’s location in on Mturk’s research studies advertisement page. Interested order to allow for the data to be downloaded. Furthermore, in participants will be asked to click a link directing them to a regard to biometric tracking data, researchers are often required series of eligibility questions. If they qualify for the study based to call upon the services of a third-party software company to on their answers, participants will complete an electronic extract the data because there currently lacks a system that informed consent and become part of the Mturk Prime panel. allows remote access to consumer devices for data extraction All panelists will receive a unique numeric participant ID. Once purposes [8]. Finally, many health-related research studies intend the panel is confirmed, all 30 panelists will complete a health to collect data from a variety of mediums including physiologic questionnaire that assesses demographics, general health, data such as heart rate and self-reported questionnaire data such physical activity, health tracking processes, and consumer as how participants view their health. The result is a technology (see Multimedia Appendix 1 for full questionnaire cumbersome data collection process that does not allow for a specifications including skip logic and response codes). smooth data acquisition and linkage process of data from varying Participants will be asked to enter their unique participant ID modes of collection. into a text box at the start of the questionnaire. At the end of the questionnaire, participants will be queried for their This study seeks to explore a protocol whereby physical activity willingness to allow researchers to download their Fitbit data. and health-related data are collected remotely through the use All varieties and models of Fitbit will be allowed in this study of personally owned activity trackers without the need for a (a range of devices is summarized in the research of Evenson face-to-face meeting with the respondents and without the use et al [13]). of study-provided devices. The primary aim of the study will be determining the feasibility of the proposed data collection Upon consent to Fitbit data access, participants will be routed protocol using an activity tracker and specifically if we are able to a third-party data service provider called Fitabase LLC (San to pair consumer wearable physiological data (ie, information Diego, California). Using the Fitbit application programming from a Fitbit activity tracker) together with self-reported interface, third-party services such as Fitabase can access and questionnaire data in order to have a better understanding of aggregate self-tracker data. Fitabase provides researchers with the health of respondents. a connection to the Fitbit infrastructure to support data collection. The research team will generate unique Fitabase Methods links for each participant. When respondents reach the end of the self-administered survey, they will click on the link to Ethics Fitabase that corresponds to their participant ID (Figure 1). Prior to initiation, this study will be reviewed and approved by Upon completion of the Fitabase sign-up, participants will be the RTI International Institutional Review Board. given $10 via Mturk Prime’s customer service team. Participants who complete the questionnaire but do not sign up with Fitabase Participant Eligibility Criteria will not receive the $10 incentive. Participants can refuse to This study will make use of a panel of Mturkers via Amazon’s participate or cancel registration at any time. We will contract Mechanical Turk (Mturk) platform. Mturkers are a workforce with Fitabase to provide the research team with 30 days of of individuals willing to participate in online research studies retrospective data and 30 days of prospective data, both from http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR RESEARCH PROTOCOLS Brinton et al the date of sign-up. After the 30 days of prospective data are fast data acquisition and the ability to easily contact participants complete, Fitabase will terminate the connection to the if there is any trouble with data collection or incentive payment. individuals’ Fitbit device. Future analyses are planned for the physiological and self-administered data to be collected throughout the study. The Study Proposed Variables following data management and data analysis plans briefly Fitabase will be used to extract daily and intraday data from the outline the proposed future acquisition, management, and linked Fitbit accounts. These variables include daily-level data analysis of these data. on total steps, distance, calories burned, total sleep time, and daily active versus sedentary time. We will also obtain Data Management hourly-level data on calories burned, active versus sedentary Study IDs will be assigned to each participant. The consumer time, heart rate, sleep, and step counts. The most granular output, device account identifiers will then be mapped to the assigned intraday data, will include minute-level step counts. These data IDs. Once the survey is complete, the responses will be exported will be downloaded as both raw and aggregated files by day, to comma-separated value (CSV) files. Separately, the consumer per person. wearable datasets will be processed and sent to the researchers from Fitabase. Both the survey responses and the consumer Study Outcomes and Data Analysis wearable dataset will be merged by ID as a de-identified, Overview compressed CSV file and formatted for analysis. This study’s main goal is to test the feasibility of extracting Data Analysis personal Fitbit data from remote survey respondents with whom Descriptive statistics will be generated for each variable. We the research team will never have direct face-to-face contact do not expect missing data to be an issue but will explore as and then linking the biometric data to self-reported questionnaire appropriate. The variability in Fitbit device type will be health data. Ease of contact, maintenance, troubleshooting, and described, as well as the type, quality, and fidelity of the data collecting participant Fitbit data throughout the study will be a collected. We will explore the Fitbit results with self-reported vital determinant of success. More specifically, the success of characteristics such as how steps per day vary by gender and the protocol will be measured by the ability to collect data from age. participants without face-to-face contact and high costs but with Figure 1. Example screen showing the Fitabase linkage. http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR RESEARCH PROTOCOLS Brinton et al adept at performing tasks with technology, thus making Results feasibility of the protocol administration more likely to be successful. Data collection was conducted between April 11, 2016, and May 12, 2016. Data analysis will take place in 2017. Second, this study will require respondents to have access to a Fitbit device in order to participate. Individuals who can afford Discussion and use Fitbit devices and other consumer wearables are more likely to be younger (between the ages of 18 and 34 years) and Summary affluent [14], thus impacting generalizability. Third, the initial This study provides a unique and innovative protocol for remote process of gathering a panel of participants will require data collection using a common physical activity tracker. The respondents to complete a screener and then at a later date, study will be cost effective and easily manageable in that complete a questionnaire in order to allow researchers to choose researchers do not need to meet with participants face-to-face a varied participant pool who all use Fitbit devices. A more at any point in the study and participants are able to use their streamlined process would be preferred in future studies own personal device to participate in the study. whereby participants would be able to fill out the screener and immediately begin the questionnaire if they are eligible, without With the acquisition of these data, we will be able to learn the need to create a panel of participants. Unfortunately, this detailed information about the health of these individuals study will not be able to employ this methodology due to panel without meeting the participant face-to-face for an interview or restrictions. in-person physiological assessments. Furthermore, we will learn more about the ease at which participants navigated the Conclusion questionnaire-to-Fitabase linkage system by determining what This study will demonstrate that activity tracker data (ie, Fitbit proportion of participants were able to complete the online data) can be remotely gathered from participants without questionnaire but were unable to connect their personal Fitbit face-to-face contact and with the use of respondent’s personal to the Fitabase platform. Finally, these data will indicate the consumer wearable devices. Future research could investigate frequency at which users sync their Fitbit, allowing us to learn the feasibility of remote data collection without the need of a more about the normal use and wearing habits of Fitbit users. 2-step data management process as well as assess the clinical validity of consumer wearable devices, like Fitbit, to ensure Limitations that the data are accurate. If effective, this methodology could This feasibility study has several limitations. First, this study be used as a guide for researchers to implement when setting is targeted to a specific population, and Mturkers may not be up a remote data collection system and could be applied to other generalizable to other populations. However, Mturkers are consumer wearable devices as well. familiar with the online environment and therefore may be more Acknowledgments RTI International would like to thank the Mturk Prime team. This project was funded from the iSHARE innovation, research, and development project. Conflicts of Interest None declared. Multimedia Appendix 1 Survey specifications for questionnaire to assess demographics, general health, physical activity, health tracking processes, and consumer technology adoption. [PDF File (Adobe PDF File), 115KB-Multimedia Appendix 1] References 1. Fitbit: Our Technology. URL: https://www.fitbit.com/technology [accessed 2017-04-08] [WebCite Cache ID 6pa1bf4p4] 2. Ortiz AM, Tueller SJ, Cook SL, Furberg RD. ActiviTeen: a protocol for deployment of a consumer wearable device in an academic setting. JMIR Res Protoc 2016 Jul;5(3):e153 [FREE Full text] [doi: 10.2196/resprot.5934] [Medline: 27457824] 3. Ahern DK, Kreslake JM, Phalen JM. What is eHealth (6): perspectives on the evolution of eHealth research. J Med Internet Res 2006 Mar;8(1):e4 [FREE Full text] [doi: 10.2196/jmir.8.1.e4] [Medline: 16585029] 4. Dobkin BH. Wearable motion sensors to continuously measure real-world physical activities. Curr Opin Neurol 2013 Dec;26(6):602-608 [FREE Full text] [doi: 10.1097/WCO.0000000000000026] [Medline: 24136126] 5. Ilies R, Dimotakis N, De Pater IE. Psychological and physiological reactions to high workloads: implications for well-being. Personnel Psychol 2010;63(2):407. [doi: 10.1111/j.1744-6570.2010.01175.x] 6. Ebner-Priemer UW, Trull TJ. Ambulatory assessment. Eur Psychol 2009 Jan;14(2):109-119. [doi: 10.1027/1016-9040.14.2.109] http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR RESEARCH PROTOCOLS Brinton et al 7. Shiffman S, Stone AA, Hufford MR. Ecological momentary assessment. Ann Rev Clin Psychol 2008 Apr;4(1):1-32. [doi: 10.1146/annurev.clinpsy.3.022806.091415] 8. Cadmus-Bertram L, Marcus BH, Patterson RE, Parker BA, Morey BL. Use of the Fitbit to measure adherence to a physical activity intervention among overweight or obese, postmenopausal women: self-monitoring trajectory during 16 weeks. JMIR Mhealth Uhealth 2015 Nov;3(4):e96 [FREE Full text] [doi: 10.2196/mhealth.4229] [Medline: 26586418] 9. Christenson D, Glick D. Crowdsourcing panel studies and real-time experiments in Mturk. Political Methodologist 2013;20(2). 10. Casler K, Bickel L, Hackett E. Separate but equal? A comparison of participants and data gathered via Amazon's MTurk, social media, and face-to-face behavioral testing. Comp Hum Behav 2013 Nov;29(6):2156-2160. [doi: 10.1016/j.chb.2013.05.009] 11. Buhrmester M, Kwang T, Gosling SD. Amazon's Mechanical Turk: a new source of inexpensive, yet high-quality, data? Perspect Psychol Sci 2011 Feb 03;6(1):3-5. [doi: 10.1177/1745691610393980] 12. Goodman J, Cryder CE, Cheema A. Data collection in a flat world: the strengths and weaknesses of Mechanical Turk samples. Behav Decis Making 2013. 13. Evenson KR, Goto MM, Furberg RD. Systematic review of the validity and reliability of consumer-wearable activity trackers. Int J Behav Nutr Phys Act 2015;12(1):159 [FREE Full text] [doi: 10.1186/s12966-015-0314-1] [Medline: 26684758] 14. Callaway J, Rozar T. Quantified wellness: wearable technology usage and market summary. RGA Reinsurance Company Abbreviations CSV: comma-separated values Mturk: Mechanical Turk Edited by G Eysenbach; submitted 19.08.16; peer-reviewed by G Dominick, M Pobiruchin; comments to author 04.10.16; revised version received 19.01.17; accepted 14.03.17; published 27.04.17 Please cite as: Brinton JE, Keating MD, Ortiz AM, Evenson KR, Furberg RD JMIR Res Protoc 2017;6(4):e66 URL: http://www.researchprotocols.org/2017/4/e66/ doi: 10.2196/resprot.6513 PMID: 28450274 ©Julia E Brinton, Mike D Keating, Alexa M Ortiz, Kelly R Evenson, Robert D Furberg. Originally published in JMIR Research Protocols (http://www.researchprotocols.org), 27.04.2017. This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included. http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 5 (page number not for citation purposes) XSL FO RenderX http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JMIR Research Protocols JMIR Publications

Establishing Linkages Between Distributed Survey Responses and Consumer Wearable Device Datasets: A Pilot Protocol

Loading next page...
 
/lp/jmir-publications/establishing-linkages-between-distributed-survey-responses-and-0p9bB7Aw0K

References (18)

Publisher
JMIR Publications
Copyright
Copyright © The Author(s). Licensed under Creative Commons Attribution cc-by 4.0
ISSN
1929-0748
DOI
10.2196/resprot.6513
Publisher site
See Article on Publisher Site

Abstract

Background: As technology increasingly becomes an integral part of everyday life, many individuals are choosing to use wearable technology such as activity trackers to monitor their daily physical activity and other health-related goals. Researchers would benefit from learning more about the health of these individuals remotely, without meeting face-to-face with participants and avoiding the high cost of providing consumer wearables to participants for the study duration. Objective: The present study seeks to develop the methods to collect data remotely and establish a linkage between self-reported survey responses and consumer wearable device biometric data, ultimately producing a de-identified and linked dataset. Establishing an effective protocol will allow for future studies of large-scale deployment and participant management. Methods: A total of 30 participants who use a Fitbit will be recruited on Mechanical Turk Prime and asked to complete a short online self-administered questionnaire. They will also be asked to connect their personal Fitbit activity tracker to an online third-party software system, called Fitabase, which will allow access to 1 month’s retrospective data and 1 month’s prospective data, both from the date of consent. Results: The protocol will be used to create and refine methods to establish linkages between remotely sourced and de-identified survey responses on health status and consumer wearable device data. Conclusions: The refinement of the protocol will inform collection and linkage of similar datasets at scale, enabling the integration of consumer wearable device data collection in cross-sectional and prospective cohort studies. (JMIR Res Protoc 2017;6(4):e66) doi: 10.2196/resprot.6513 KEYWORDS Fitbit; Mturk; mHealth; clinical research protocol; consumer wearable; physical activity tracker quality and duration of sleep, heart rate, and location [1]. Introduction Researchers would benefit from being able to remotely gather and link these data without the need for face-to-face interaction, The increasing variety, functionality, and storage capacity of encouraging the use of the respondent’s own devices in the data consumer wearable devices has created an opportunity for using collection process. the data collected on these devices for research purposes. Consumer wearables are most commonly worn on the wrist but A remote data collection protocol would reduce the cost of may also be worn on clothing, at the waist, or as part of eyewear providing devices to participants, reduce the time spent meeting or earwear. In this paper, we focus on activity trackers, a type and training respondents on their use, and increase the speed at of consumer wearable which can collect a variety of data which data could be collected [2]. Furthermore, with the ability including steps, distance, physical activity, calories burned, to efficiently collect health data remotely, hospitals, worksites, http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR RESEARCH PROTOCOLS Brinton et al and other health care providers could monitor participants in in exchange for money deposited into their Amazon.com real time, reducing the need for frequent check-ups and the personal account. With the help of Mturk Prime, this study will financial strain that is associated with those appointments [3]. use Mturkers as a basis for the sample. Mturk Prime operates The utility and success of remote-access interventions was as a team of online support staff who assist researchers in demonstrated to be effective in collecting biometric (eg, managing, communicating with, and collecting data from continuous heart rate, sleep, and other health indicators) [4] Mturkers (www.turkprime.com). Research using Mturkers has data, yet research using consumer wearable devices is limited increased dramatically in the past few years due to the low cost and presents challenges. and vast acquisition of data [9]. Additionally, studies show that Mturkers are more demographically diverse than standard While research investigating the viability and functionality of convenience samples and samples from other online forms of consumer wearables has increased in recent years, thanks in data collection such as Twitter [10] and may be generalizable part to calls for research from the National Institutes of Health to the greater population [11,12]. In order for an Mturker to and United Nations International Children's Fund, challenges become a part of this study’s panel, the person will be required still exist in the utility of these devices as the most efficient and to either keep track of their own weight, diet, or exercise routine useful way to learn about the health of an individual. Frequently, or keep track of their own blood pressure, blood sugar, sleep remote data collection may be hindered by burdensome patterns, headaches, or some other health-related indicator. An notification systems, forcing individuals to use study-provided eligible Mturker must also be at least 18 years of age, regularly devices they are not comfortable with and requiring frequent wear a Fitbit, and be willing to give the research team access face-to-face contact with participants in order to download data to their Fitbit data for the previous month and the upcoming from their devices. For example, many present studies employ month from the date of sign-up. experience sampling methodology, which requires that researchers notify participants throughout the day requesting Study Design and Procedure that they provide their remote data, a burdensome process for The sample will consist of 30 Mturk participants. Participants both respondents and researchers [5,6,7]. They often frequently will read a short task description and compensation information require that participants travel to the researcher’s location in on Mturk’s research studies advertisement page. Interested order to allow for the data to be downloaded. Furthermore, in participants will be asked to click a link directing them to a regard to biometric tracking data, researchers are often required series of eligibility questions. If they qualify for the study based to call upon the services of a third-party software company to on their answers, participants will complete an electronic extract the data because there currently lacks a system that informed consent and become part of the Mturk Prime panel. allows remote access to consumer devices for data extraction All panelists will receive a unique numeric participant ID. Once purposes [8]. Finally, many health-related research studies intend the panel is confirmed, all 30 panelists will complete a health to collect data from a variety of mediums including physiologic questionnaire that assesses demographics, general health, data such as heart rate and self-reported questionnaire data such physical activity, health tracking processes, and consumer as how participants view their health. The result is a technology (see Multimedia Appendix 1 for full questionnaire cumbersome data collection process that does not allow for a specifications including skip logic and response codes). smooth data acquisition and linkage process of data from varying Participants will be asked to enter their unique participant ID modes of collection. into a text box at the start of the questionnaire. At the end of the questionnaire, participants will be queried for their This study seeks to explore a protocol whereby physical activity willingness to allow researchers to download their Fitbit data. and health-related data are collected remotely through the use All varieties and models of Fitbit will be allowed in this study of personally owned activity trackers without the need for a (a range of devices is summarized in the research of Evenson face-to-face meeting with the respondents and without the use et al [13]). of study-provided devices. The primary aim of the study will be determining the feasibility of the proposed data collection Upon consent to Fitbit data access, participants will be routed protocol using an activity tracker and specifically if we are able to a third-party data service provider called Fitabase LLC (San to pair consumer wearable physiological data (ie, information Diego, California). Using the Fitbit application programming from a Fitbit activity tracker) together with self-reported interface, third-party services such as Fitabase can access and questionnaire data in order to have a better understanding of aggregate self-tracker data. Fitabase provides researchers with the health of respondents. a connection to the Fitbit infrastructure to support data collection. The research team will generate unique Fitabase Methods links for each participant. When respondents reach the end of the self-administered survey, they will click on the link to Ethics Fitabase that corresponds to their participant ID (Figure 1). Prior to initiation, this study will be reviewed and approved by Upon completion of the Fitabase sign-up, participants will be the RTI International Institutional Review Board. given $10 via Mturk Prime’s customer service team. Participants who complete the questionnaire but do not sign up with Fitabase Participant Eligibility Criteria will not receive the $10 incentive. Participants can refuse to This study will make use of a panel of Mturkers via Amazon’s participate or cancel registration at any time. We will contract Mechanical Turk (Mturk) platform. Mturkers are a workforce with Fitabase to provide the research team with 30 days of of individuals willing to participate in online research studies retrospective data and 30 days of prospective data, both from http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR RESEARCH PROTOCOLS Brinton et al the date of sign-up. After the 30 days of prospective data are fast data acquisition and the ability to easily contact participants complete, Fitabase will terminate the connection to the if there is any trouble with data collection or incentive payment. individuals’ Fitbit device. Future analyses are planned for the physiological and self-administered data to be collected throughout the study. The Study Proposed Variables following data management and data analysis plans briefly Fitabase will be used to extract daily and intraday data from the outline the proposed future acquisition, management, and linked Fitbit accounts. These variables include daily-level data analysis of these data. on total steps, distance, calories burned, total sleep time, and daily active versus sedentary time. We will also obtain Data Management hourly-level data on calories burned, active versus sedentary Study IDs will be assigned to each participant. The consumer time, heart rate, sleep, and step counts. The most granular output, device account identifiers will then be mapped to the assigned intraday data, will include minute-level step counts. These data IDs. Once the survey is complete, the responses will be exported will be downloaded as both raw and aggregated files by day, to comma-separated value (CSV) files. Separately, the consumer per person. wearable datasets will be processed and sent to the researchers from Fitabase. Both the survey responses and the consumer Study Outcomes and Data Analysis wearable dataset will be merged by ID as a de-identified, Overview compressed CSV file and formatted for analysis. This study’s main goal is to test the feasibility of extracting Data Analysis personal Fitbit data from remote survey respondents with whom Descriptive statistics will be generated for each variable. We the research team will never have direct face-to-face contact do not expect missing data to be an issue but will explore as and then linking the biometric data to self-reported questionnaire appropriate. The variability in Fitbit device type will be health data. Ease of contact, maintenance, troubleshooting, and described, as well as the type, quality, and fidelity of the data collecting participant Fitbit data throughout the study will be a collected. We will explore the Fitbit results with self-reported vital determinant of success. More specifically, the success of characteristics such as how steps per day vary by gender and the protocol will be measured by the ability to collect data from age. participants without face-to-face contact and high costs but with Figure 1. Example screen showing the Fitabase linkage. http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR RESEARCH PROTOCOLS Brinton et al adept at performing tasks with technology, thus making Results feasibility of the protocol administration more likely to be successful. Data collection was conducted between April 11, 2016, and May 12, 2016. Data analysis will take place in 2017. Second, this study will require respondents to have access to a Fitbit device in order to participate. Individuals who can afford Discussion and use Fitbit devices and other consumer wearables are more likely to be younger (between the ages of 18 and 34 years) and Summary affluent [14], thus impacting generalizability. Third, the initial This study provides a unique and innovative protocol for remote process of gathering a panel of participants will require data collection using a common physical activity tracker. The respondents to complete a screener and then at a later date, study will be cost effective and easily manageable in that complete a questionnaire in order to allow researchers to choose researchers do not need to meet with participants face-to-face a varied participant pool who all use Fitbit devices. A more at any point in the study and participants are able to use their streamlined process would be preferred in future studies own personal device to participate in the study. whereby participants would be able to fill out the screener and immediately begin the questionnaire if they are eligible, without With the acquisition of these data, we will be able to learn the need to create a panel of participants. Unfortunately, this detailed information about the health of these individuals study will not be able to employ this methodology due to panel without meeting the participant face-to-face for an interview or restrictions. in-person physiological assessments. Furthermore, we will learn more about the ease at which participants navigated the Conclusion questionnaire-to-Fitabase linkage system by determining what This study will demonstrate that activity tracker data (ie, Fitbit proportion of participants were able to complete the online data) can be remotely gathered from participants without questionnaire but were unable to connect their personal Fitbit face-to-face contact and with the use of respondent’s personal to the Fitabase platform. Finally, these data will indicate the consumer wearable devices. Future research could investigate frequency at which users sync their Fitbit, allowing us to learn the feasibility of remote data collection without the need of a more about the normal use and wearing habits of Fitbit users. 2-step data management process as well as assess the clinical validity of consumer wearable devices, like Fitbit, to ensure Limitations that the data are accurate. If effective, this methodology could This feasibility study has several limitations. First, this study be used as a guide for researchers to implement when setting is targeted to a specific population, and Mturkers may not be up a remote data collection system and could be applied to other generalizable to other populations. However, Mturkers are consumer wearable devices as well. familiar with the online environment and therefore may be more Acknowledgments RTI International would like to thank the Mturk Prime team. This project was funded from the iSHARE innovation, research, and development project. Conflicts of Interest None declared. Multimedia Appendix 1 Survey specifications for questionnaire to assess demographics, general health, physical activity, health tracking processes, and consumer technology adoption. [PDF File (Adobe PDF File), 115KB-Multimedia Appendix 1] References 1. Fitbit: Our Technology. URL: https://www.fitbit.com/technology [accessed 2017-04-08] [WebCite Cache ID 6pa1bf4p4] 2. Ortiz AM, Tueller SJ, Cook SL, Furberg RD. ActiviTeen: a protocol for deployment of a consumer wearable device in an academic setting. JMIR Res Protoc 2016 Jul;5(3):e153 [FREE Full text] [doi: 10.2196/resprot.5934] [Medline: 27457824] 3. Ahern DK, Kreslake JM, Phalen JM. What is eHealth (6): perspectives on the evolution of eHealth research. J Med Internet Res 2006 Mar;8(1):e4 [FREE Full text] [doi: 10.2196/jmir.8.1.e4] [Medline: 16585029] 4. Dobkin BH. Wearable motion sensors to continuously measure real-world physical activities. Curr Opin Neurol 2013 Dec;26(6):602-608 [FREE Full text] [doi: 10.1097/WCO.0000000000000026] [Medline: 24136126] 5. Ilies R, Dimotakis N, De Pater IE. Psychological and physiological reactions to high workloads: implications for well-being. Personnel Psychol 2010;63(2):407. [doi: 10.1111/j.1744-6570.2010.01175.x] 6. Ebner-Priemer UW, Trull TJ. Ambulatory assessment. Eur Psychol 2009 Jan;14(2):109-119. [doi: 10.1027/1016-9040.14.2.109] http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR RESEARCH PROTOCOLS Brinton et al 7. Shiffman S, Stone AA, Hufford MR. Ecological momentary assessment. Ann Rev Clin Psychol 2008 Apr;4(1):1-32. [doi: 10.1146/annurev.clinpsy.3.022806.091415] 8. Cadmus-Bertram L, Marcus BH, Patterson RE, Parker BA, Morey BL. Use of the Fitbit to measure adherence to a physical activity intervention among overweight or obese, postmenopausal women: self-monitoring trajectory during 16 weeks. JMIR Mhealth Uhealth 2015 Nov;3(4):e96 [FREE Full text] [doi: 10.2196/mhealth.4229] [Medline: 26586418] 9. Christenson D, Glick D. Crowdsourcing panel studies and real-time experiments in Mturk. Political Methodologist 2013;20(2). 10. Casler K, Bickel L, Hackett E. Separate but equal? A comparison of participants and data gathered via Amazon's MTurk, social media, and face-to-face behavioral testing. Comp Hum Behav 2013 Nov;29(6):2156-2160. [doi: 10.1016/j.chb.2013.05.009] 11. Buhrmester M, Kwang T, Gosling SD. Amazon's Mechanical Turk: a new source of inexpensive, yet high-quality, data? Perspect Psychol Sci 2011 Feb 03;6(1):3-5. [doi: 10.1177/1745691610393980] 12. Goodman J, Cryder CE, Cheema A. Data collection in a flat world: the strengths and weaknesses of Mechanical Turk samples. Behav Decis Making 2013. 13. Evenson KR, Goto MM, Furberg RD. Systematic review of the validity and reliability of consumer-wearable activity trackers. Int J Behav Nutr Phys Act 2015;12(1):159 [FREE Full text] [doi: 10.1186/s12966-015-0314-1] [Medline: 26684758] 14. Callaway J, Rozar T. Quantified wellness: wearable technology usage and market summary. RGA Reinsurance Company Abbreviations CSV: comma-separated values Mturk: Mechanical Turk Edited by G Eysenbach; submitted 19.08.16; peer-reviewed by G Dominick, M Pobiruchin; comments to author 04.10.16; revised version received 19.01.17; accepted 14.03.17; published 27.04.17 Please cite as: Brinton JE, Keating MD, Ortiz AM, Evenson KR, Furberg RD JMIR Res Protoc 2017;6(4):e66 URL: http://www.researchprotocols.org/2017/4/e66/ doi: 10.2196/resprot.6513 PMID: 28450274 ©Julia E Brinton, Mike D Keating, Alexa M Ortiz, Kelly R Evenson, Robert D Furberg. Originally published in JMIR Research Protocols (http://www.researchprotocols.org), 27.04.2017. This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included. http://www.researchprotocols.org/2017/4/e66/ JMIR Res Protoc 2017 | vol. 6 | iss. 4 | e66 | p. 5 (page number not for citation purposes) XSL FO RenderX

Journal

JMIR Research ProtocolsJMIR Publications

Published: Apr 27, 2017

Keywords: Fitbit; Mturk; mHealth; clinical research protocol; consumer wearable; physical activity tracker

There are no references for this article.