Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/unpaywall/why-don-t-men-ever-stop-to-ask-for-directions-gender-social-influence-qS1Y7hydJD
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- Unpaywall
- ISSN
- 0276-7783
- DOI
- 10.2307/3250981
- Publisher site
- See Article on Publisher Site
Abstract
the integration of subjective norm into the model. In light of these findings, implications for theory Using the Technology Acceptance Model (TAM), and practice are discussed. this research investigated gender differences in the overlooked context of individual adoption and Keywords: User acceptance, adoption, techno- sustained usage of technology in the workplace. logy acceptance model, social influences, gender User reactions and technology usage behavior differences were studied over a five-month period among 342 workers being introduced to a new software ISRL Categories: AA01, AA07, AC0401, AI0108 system. At all three points of measurement, com- pared to women, men's technology usage deci- Introduction While advances in hardware and software capa- Ron Weber was the accepting senior editor for this bilities continue at an unprecedented pace, the paper. MIS Quarterly Vol. 24 No. 1, pp. 115-139/March 2000 115 Venkatesh & Morris/Gender in Technology Acceptance and Usage problem of underutilized systems remains (Johan- Interestingly, TAM’s referent theory (i.e., TRA) sen and Swigart 1996; Moore 1991; Norman includes social influence via a construct called 1993; Weiner 1993). Importantly, low usage has subjective norm. Much prior research in psych- been listed as one of the underlying causes be- ology (see Ajzen 1991 for a review) found hind the so-called “productivity paradox” (Lan- subjective norm to be an important determinant of dauer 1995; Sichel 1997). Understanding the intention and/or behavior. However, TAM ex- conditions under which information systems are or cluded this construct due to theoretical and are not accepted and used within organizations measurement problems (see Davis et al. 1989). continues to be an important issue. Information Although subjective norm can be expected to be systems research has examined user acceptance important in determining technology acceptance and usage behavior from several different and usage based on TRA and the Theory of perspectives. Among the different models that Planned Behavior (TPB) (Ajzen 1985, 1991), have been proposed, the Technology Acceptance empirical evidence supporting the role of the Model (TAM) (Davis 1989; Davis et al. 1989), construct has been somewhat mixed. Some adapted from the Theory of Reasoned Action investigations have omitted the construct com- (TRA) (Ajzen and Fishbein 1980; Fishbein and pletely (e.g., Adams et al. 1992; Szajna 1994, Ajzen 1975), offers a powerful and parsimonious 1996). Others have found the construct to be explanation for user acceptance and usage non-significant (e.g., Davis et al. 1989; Mathieson behavior. TAM posits that user acceptance is 1991). Still others have found the construct to be determined by two key beliefs, namely perceived significant (e.g., Hartwick and Barki 1994; Taylor usefulness and perceived ease of use. Perceived and Todd 1995b). Nonetheless, given that other usefulness (U) is defined as the extent to which a theoretical perspectives emphasize the impor- person believes that using a particular technology tance of social aspects of technology use will enhance her/his job performance, while including critical mass (Markus 1990), social perceived ease of use (EOU) is defined as the influence (Fulk et al. 1987), adaptive structuration degree to which a person believes that using a (Poole and DeSanctis 1990), hermeneutic technology will be free from effort (Davis 1989). interpretation (Lee 1994), and critical social theory The robustness of TAM has been established (Ngwenyama and Lee 1997), we believe it is through several applications and replications important to investigate whether social influence (Adams et al. 1992; Chin and Todd 1995; Davis should be integrated into TAM. Since the de- 1989, 1993; Davis et al. 1989; Davis and velopment of TAM, even within the context of Venkatesh 1996; Gefen and Straub 1997; Igbaria rational perspectives (e.g., TRA, TPB, and TAM), et al. 1997; Mathieson 1991; Morris and Dillon recent research has successfully operationalized 1997; Segars and Grover 1993; Subramanian subjective norm (see Mathieson 1991; Taylor and 1994; Szajna 1994, 1996; Taylor and Todd 1995b; Todd 1995a, 1995b). Venkatesh 1999; Venkatesh and Davis 1996). Perhaps surprisingly, gender’s role within TAM has been investigated only recently (Gefen and Two important constructs that have received very Straub 1997). So far, however, research has little attention in the context of TAM research are studied only gender-based perceptual differences social influence and gender (cf. Gefen and Straub and not gender-based differences in decision 1997). These two constructs are potentially criti- making processes about technology. Nonethe- cal to our understanding of user acceptance since less, psychology research that has studied gender they could both play an important role in deter- differences in decision making processes indi- mining how users make their decisions about cates that schematic processing by women and adopting and using new technologies. In fact, men is different (cf. Bem and Allen 1974). For there is a significant body of evidence outside the instance, from an information processing perspec- domain of information systems in general sup- tive, there are known differences in determinants porting the viewpoint that social influence and of self-esteem between both sexes (Tashakkori 1993). Such a view is consistent with Bem (1981), gender do indeed play a critical role in influencing who argues that women and men encode and behaviors in a wide variety of domains. 116 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage process information using different socially- as user experience increases (e.g., Davis et al. constructed cognitive structures that, in turn, help 1989). Therefore, to help gain a thorough under- determine and direct an individual’s perceptions. standing of the underlying phenomena, this As a result, individuals tend to make decisions research studies the role of gender in initial which reflect biases inherent in the individual’s technology acceptance decisions and continued perceptions and actions (e.g., Nisbett and Ross usage behavior decisions. The moderating role of 1980). This means that gender schemas can be gender is expected to continue with increasing considered to be a normative guide (Kagan 1964; user experience (with the target system) with one Kohlberg 1966) that causes unconscious or exception: subjective norm is not expected to be internalized action consistent with the schema. a significant determinant of intention with in- creasing experience for women or men. Given these important missing elements in TAM research, in this paper, we describe research that seeks to extend TAM to include subjective norm Short-Term Effects and gender. Specifically, taking a longitudinal ap- proach with data gathered from five organizations, Perceived Usefulness we seek to achieve three primary objectives: Perceived usefulness (U) is defined as the extent to which a person believes that using a particular 1. Understand gender differences in the relative technology will enhance her/his job performance influence of the original TAM constructs (Davis 1989). Perceived usefulness, which re- (perceived usefulness and perceived ease of flects perceptions of the performance-use con- use) on intention to use a new technology. tingency, has been closely linked to outcome expectations, instrumentality, and extrinsic motiva- 2. Integrate subjective norm into TAM using tion (Davis 1989, 1993; Davis et al. 1989, 1992). gender as a moderator. A significant body of TAM research has shown that perceived usefulness is a strong determinant 3. Understand gender differences over the long of user acceptance, adoption, and usage behavior term as it relates to sustained usage of (e.g., Davis 1989; Davis et al. 1989; Mathieson technology with increasing experience. 1991; Taylor and Todd 1995a, 1995b; Venkatesh and Davis forthcoming). Theoretical Development In understanding gender differences in the role of perceived usefulness as a determinant of techno- Figure 1 shows TAM, as developed by Davis et al. logy acceptance, we draw from research on (1989), together with the extensions proposed in gender differences in the salience of outcomes as this paper. Specifically, we propose that gender determinants of behavior. Prior research has indi- will moderate the perceived usefulness-intention, cated that men’s work role is typically their most perceived ease of use-intention, subjective norm- salient, while the family role is often only of intention, and perceived ease of use-perceived secondary importance (e.g., Barnett and Marshall usefulness relationships. W e further examine the 1991). For example, O’Neill (1982) suggests that role of experience as an additional moderator of men may place great emphasis on work, accom- the different relationships. In studying acceptance plishment, and eminence. Hoffman (1972) points and use of a technology, it is important to examine out that men are motivated by achievement needs the phenomenon over a duration of time since to a greater extent than women. These argu- users will evolve from being novices to ex- ments suggest that men, more than women, are perienced users of the new system (e.g., Davis et directed toward individualistic tasks and goals al. 1989). This is of particular importance since (Carlson 1971; Gill et al. 1987; see also Stein and during the earliest stages of technology intro- Bailey 1973). Other gender-related differences duction, users are making an “acceptance” deci- have also been reported in the literature. For sion, which has been shown to differ systema- example, some researchers have shown that tically from “usage” decisions over the long term male-valued traits include “objective” and “logical” MIS Quarterly Vol. 24 No. 1/March 2000 117 Venkatesh & Morris/Gender in Technology Acceptance and Usage Per Perc cei eiv ved ed U Us seful efulne nes ss s H1 H1,, H4 H4 H H2b, 2b, H H5 5b b H H2a, 2a, H5 H5a a Per Perc cei eiv ved ed Be Behav haviio or ra all Behav Behaviio or r Eas Ease e of of U Us se e In Inte ten nttiio on n H3, H3, H6 H6 Subj Subjec ecttiiv ve e No Nor rm m G Ge ender nder Ex Exper periienc ence e T Te ec ch hnol nolo og gy y A Ac cc ceptanc eptance e M M ode odell:: F Fiinal nal M M o od de ell ( (D Dav aviis s et et al al.. 1989) 1989) Figure 1. Technology Acceptance Model: Proposed Extensions (Rosenkrantz et al. 1968). Furthermore, as opera- Perceived Ease of Use Perceived ease of use (EOU) is defined as the tionalized by Bem’s Sex Role Inventory (BSRI) degree to which a person believes that using the (Bem 1981), men tend to exhibit more “masculine” system will be free from effort (Davis 1989). traits (e.g., assertiveness) compared to women. Perceived ease of use has been shown to have Meta-analytic evidence by Taylor and Hall (1982) an effect on intention via two causal pathways: (1) indicates that masculine scales are highly cor- a direct effect on intention and (2) an indirect related with instrumental behaviors. Hofstede’s effect on intention via perceived usefulness (EOU- (1980) seminal work on culture found that men U-BI). The direct effect suggests that perceived rate the potential for advancement and earning ease of use could be a potential catalyst to power—two classic extrinsic motivators—as more increasing the likelihood of user acceptance. The important than women. Given the weight of the indirect effect is explained as stemming from a evidence, Minton and Schneider (1980) conclude situation where, other things being equal, the that men may be more task oriented than women. easier a technology is to use, the more useful it In this context, task orientation refers to the ac- can be (Davis et al.(1989). With little or no prior complishment of organizational tasks that may experience, prior research has demonstrated that require technology use. Therefore, we expect the direct causal pathway (i.e., EOU-BI) is most factors that are related to productivity enhance- relevant, and the indirect effect via perceived ment to be more salient for men. usefulness is somewhat less important (see Davis et al. 1989; Szajna 1996). To understand gender H1: Perceived usefulness will influence differences in the role of perceived ease of use, behavioral intention to use a system therefore, we must understand differences in both the direct and indirect effects of perceived ease of more strongly for men than it will use on intention. influence women. 118 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage Beginning with the theoretical development and H2a: Perceived ease of use will operationalization of the construct, perceived ease influence behavioral intention to use a of use has been closely related to self-efficacy system more strongly for women than it (Bandura 1977, 1982, 1986). There is much will influence men. evidence in psychology (Chan and Fishbein 1993; Sparks 1994; see also Fishbein and Stasson As proposed in H1, men appear highly motivated 1990) and information systems (Venkatesh by productivity-related factors like usefulness forthcoming; Venkatesh and Davis 1996) sup- (Minton and Schneider 1980). Davis et al. (1989) porting computer self-efficacy (one’s judgment showed that perceived ease of use is a deter- about one’s ability to use a computer for a specific minant of perceived usefulness. They interpret task) as a determinant of perceptions of ease/ this relationship by stating that systems that are difficulty. In the context of technology acceptance easier to use may ultimately be more useful. and usage in the workplace, evidence indicates Thus, systems that are perceived as easier to use that providing support staff is a very important will facilitate system use and task accomplishment organizational response to help users overcome more than systems that are seen as difficult to barriers and hurdles to technology use especially use. In other words, the system that is easier to during the early stages of learning and use (e.g., use will generate the best cost/benefit ratio for Bergeron et al. 1990). This is consistent with achievement-oriented individuals. For example, Hofstede’s contention that women rate the impor- users of modern personal computers generally tance of service aspects and physical environment consider graphical user interfaces to be more more highly than men. Therefore, we expect per- productive than older text-based interfaces ceived ease of use to be more salient for women because they are easier to use—although when compared to its salience for men. objectively, they may not be more “useful” than the older style interface. It seems that individuals for whom task achievement is most salient would There is additional theoretical justification sup- be influenced more significantly by perceived ease porting such an effect. Women typically display of use. lower computer aptitude (Felter 1985) and higher levels of computer anxiety (Morrow et al. 1986; H2b: Perceived ease of use will see Rosen and Maguire 1990 for a review) influence perceived usefulness more compared to men. IS research also supports the strongly for men than it will influence existence of higher levels of computer anxiety women. among women (e.g., Igbaria and Chakrabarti 1990). Further, there is recent evidence from real- world settings that women tend to be more Subjective Norm anxious than men about computer use (Bozio- Subjective norm (SN) is defined as the degree to nelos 1996). A significant body of research in which an individual believes that people who are psychology (e.g., Hunt and Bohlin 1993) has important to her/him think she/he should perform shown an inverse relationship between computer the behavior in question (Fishbein and Ajzen anxiety and computer self-efficacy, a known deter- 1975). In the technology domain, both peer and minant of perceived ease of use (Venkatesh and superior influences have been shown to be strong Davis 1996). Thus, given the intertwining of determinants of subjective norm (Mathieson 1991; anxiety and self-efficacy, higher levels of compu- Taylor and Todd 1995b). Therefore, in examining gender differences in subjective norm, it is useful ter anxiety among women can be expected to lead to understand the degree to which women/men to lowering of self-efficacy, which in turn could can be influenced and the extent to which they lead to lowering of ease of use perceptions. Since respond to information provided by other perceived ease of use has typically been seen as referents. a hurdle to user acceptance (Venkatesh and Davis 1996), low evaluations of ease of use can As implied earlier, women exhibit more “feminine” cause an increase in the salience of such percep- traits (e.g., tenderness), as operationalized by the tions in determining user acceptance decisions. MIS Quarterly Vol. 24 No. 1/March 2000 119 Venkatesh & Morris/Gender in Technology Acceptance and Usage BSRI (Bem 1981). Meta-analytic evidence also A separate and distinct body of research has examined differences in susceptibility to influence suggests that women are more “expressive” but has suggested an alternative causal compared to men (Taylor and Hall 1982). mechanism. For example, evidence suggests that Additional evidence indicates that women are women are more attentive to social cues in the strongly motivated by affiliation needs (Hoffman environment while men attend to other stimuli 1972) and prefer person-oriented professions such as objects and/or visual patterns (e.g., Garai (Weller et al. 1976). Consistent with this view, and Scheinfeld 1968; Parsons and Bales 1955; other studies show that women are more disposed W illiams and Best 1982). Others have suggested toward interpersonal goals and success in inter- that women and men are equally attentive to personal relationships (see Carlson 1971; Gill et social cues in the environment (e.g., Roberts al. 1987; Stein and Bailey 1973). This outcome 1991); however, women are more responsive to may be attributed to women having a greater those cues (i.e., they yield more to social awareness of others’ feelings compared to men pressures). Roberts suggests that this may be (Rosenkrantz et al. 1968). In related research, because men adopt a competitive, potentially Skitka and Maslach (1996) reported that women overconfident attitude (see Lundeberg et al. 1994) used constructs more related with the harmonious about others’ evaluations, while women are more functioning of groups, interrelationships, and con- accepting of others’ opinions. This suggests that women may look at others’ opinions as cern with the overall “communion” of the group in opportunities to learn more about their own the process of describing others. Within the abilities. This line of reasoning implies that organizational environment, Landau and Leven- women may weight the opinion of other people in thal (1976) found that women were more likely to considering new technology and may factor those retain less productive employees for social opinions into the overall decision-making process reasons compared to men. Overall, women tend about adopting that technology more than men. to rate the importance of pleasing others more Although the context of investigation in prior highly than men (e.g., Miller 1986). research was not technology acceptance and use, we expect that the importance of social factors Research dating back over a decade suggests and increased deference to others’ opinions will that women and men also differ in the extent to generalize to the context of decisions about which they can be influenced by others (Becker technology and manifest itself in normative 1986; Eagly and Carli 1981). For example, pressures being more salient for women. research shows that women tend to be more compliant while men are more likely to rebel H3: Subjective norm will influence behavioral intention to use a system against requests or orders from others (e.g., more strongly for women than it will Minton et al. 1971). Similarly, women appear influence men. more likely to conform with majority opinions (Eagly 1978; Maccoby and Jacklin 1974). Based on their extensive review of the literature, Minton and Schneider concluded that women were more Long-Term Effects people-oriented while men tend to be somewhat more independent and self-confident. Due to Perceived Usefulness different socialization patterns of women in today’s Prior research on TAM provides valuable insight society compared to two decades ago, it is into the role of perceived usefulness and possible to argue that some of the findings about perceived ease of use over time with increasing women being more susceptible to influence than direct experience. A significant body of research men may be dated. Nonetheless, even recent supports the role of perceived usefulness as a evidence is consistent with a gender schema view strong determinant of user intentions and usage that women tend to be more compliant (e.g., behavior over time. For example, Davis et al. Crawford et al. 1995). (1989) found that the perceived usefulness- inten- 120 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage tion relationship remained strong over 14 weeks of 1996; Morrow et al. 1986; see Rosen and Maguire 1990 for a review) and lower computer aptitude use across multiple systems. More recently, (Felter 1985) among women that may necessitate longitudinal studies by Taylor and Todd (1995b) tapping into support staff during the early stages (12 weeks), Szajna (1996) (15 weeks), and of learning/experience and practice. Another Venkatesh and Davis (1996) (five weeks) all found potential reason for the higher salience of that perceived usefulness remains a significant perceived ease of use to women is based on the determinant of behavioral intention over time. notion that support staff will be more important to Related psychology research also supports the women than men from a social/affiliation perspec- notion that attitudinal components (such as tive. Following their early interactions with support perceived usefulness and perceived ease of use) staff in the context of the new technology, the tend to be strong determinants of intention and influence of perceived ease of use on women's behavior with increasing direct experience with the technology usage can be expected to be target behavior (Doll and Ajzen 1992; Fazio and additionally motivated from the standpoint of Zanna 1978a, 1978b, 1981; Regan and Fazio social/affiliation needs and interpersonal 1977) for up to a year (Reinecke et al. 1996). interaction. Thus, it is clear that instrumental factors (such as perceived usefulness) are not simply important H5a: With increasing direct experience with the technology, perceived ease of initial determinants of intention: they remain use will influence behavioral intention to important over the long term. Given that task- use a system more strongly for women oriented factors are more important for men than than it will influence men. for women (e.g., Minton and Schneider 1980) on an ongoing basis, we expect that gender dif- Research has shown that while the direct effects ferences in the salience of instrumental factors of perceived ease of use remain important, over that were present at the time of the initial time, the indirect effect of perceived ease of use acceptance decision will be sustained over time (through perceived usefulness) becomes stronger. with increasing direct technology experience. Therefore, given the greater achievement orien- tation for men in the long run (see H4), factors that H4: With increasing direct experience are seen as facilitating or inhibiting task accom- with the technology, perceived useful- plishment (i.e., the EOU-U link) are likely to be ness will influence behavioral intention to weighed more strongly by men as direct use a system more strongly for men than experience with the target system increases. it will influence women. Thus, we propose that while the direct influence of perceived ease of use on intention is more salient for women (see H5a), because the indirect effects Perceived Ease of Use operate through instrumental factors (U), the Recall that two causal pathways (EOU-BI, EOU-U- indirect effects of perceived ease of use on inten- BI) are important in determining user intentions. tion (via perceived usefulness) will be more Recent research has found that even with strongly weighted by men. increasing experience, both pathways remain significant (Venkatesh forthcoming; Venkatesh H5b: With increasing direct experience and Davis 1996). Prior research (e.g., Bergeron with the technology, perceived ease of et al. 1990) indicates that providing support staff use will influence perceived usefulness is a crucial element in alleviating constraints to more strongly for men than it will technology usage. As with the short-term impact influence women. of perceived ease of use, in the long run also, we expect that perceived ease of use, driven by availability of support staff to alleviate constraints Subjective Norm to technology use, will be more salient to women To understand gender differences in subjective compared to men. This is further corroborated by norm over the long term, it is necessary to con- the higher levels of computer anxiety (Bozionelos sider the role of experience and how that MIS Quarterly Vol. 24 No. 1/March 2000 121 Venkatesh & Morris/Gender in Technology Acceptance and Usage experience can influence the importance of others’ psychology has shown that the direct effect of opinions in determining intentions for any one indi- subjective norm on intention is strong in the early vidual. In the short term, we proposed that stages of a new behavior and tends to wear off women will weight the opinions of others’ more over time (e.g., Reinecke et al. 1996). In the highly than men (see H3). Others’ opinions can context of technology acceptance in voluntary be expected to be critical in the short-term when usage settings, this suggests that the influence of one has little or no prior experience with a specific peers and superiors will diminish to non- technology (i.e., in the early stages of acceptance significance over time with increasing experience and usage). Even though the contexts of tech- with the target system. nology usage examined in this research are voluntary usage contexts, normative pressures H6: With increasing direct experience from peers, superiors, IS staff, etc. can nonethe- with the technology, subjective norm will less play an important role in determining indivi- not influence behavioral intention to use dual intentions and behavior. In the early stages a system for either women or men. of user experience where user interaction with the target system has been somewhat limited, even if In sum, the current research proposes important an individual does not have a favorable reaction to extensions to the Technology Acceptance Model the system, the individual will tend to comply with using gender as a potential moderator. The others’ views and intend/use the target system to hypotheses proposed deal with gender differences attain a favorable reaction from important in roles of perceived usefulness and perceived referents. Such an effect of subjective norm on ease of use as determinants of technology intention is referred to as “compliance” (Warshaw acceptance and usage. In addition, the current 1980). work attempts to integrate subjective norm into TAM by taking a gender-oriented approach. As direct experience with technology increases Table 1 summarizes the hypotheses. over time, individuals have a better assessment of the benefits and costs associated with using that technology. Even if their original decision was Method based on others’ opinions, individuals begin to “internalize” others’ opinions especially if they are Participants and Systems consistent with the results of their own direct experience. Thus, the direct effect of subjective A total of 445 individuals from five organizations norm on behavioral intention is reduced (Oliver agreed to participate in the study. Consistent with and Bearden 1985; Warshaw 1980). The shifting the original development and purpose of TAM, all causal mechanism (i.e., from compliance to inter- participants were in the process of being intro- nalization) operative with increasing experience duced to a new technology, use of which was can also be justified from an anchoring and voluntary within the organization. In each organi- adjustment perspective from behavioral decision zation, the new technology being introduced could be broadly classified as a system for data and theory. A significant body of research (e.g., information retrieval. Although the specific system Bettman and Sujan 1987; Mervis and Rosch introduced in each organization was different, the 1980), including recent IS research (Venkatesh general characteristics of the technology intro- forthcoming), has suggested that in the absence duction and usage processes (e.g., training, of direct behavioral experience with the target voluntariness of use) were comparable. Concep- object, individuals anchor their perceptions to tually, these were considered important to permit general/ abstract criteria, which in this case the pooling of data across technologies/organi- includes complying with the ideas of peers and zations. (Note: We discuss the statistical analy- superiors. With increasing experience, user sis issues related to pooling in the results section.) judgments reflect specific/ concrete criteria that Pooling data across different technologies/organi- result from the interaction with the target object zations is consistent with prior research in user (i.e., new system) and less from normative in- acceptance (e.g., Compeau and Higgins 1995b; fluences. Consistent with this view, research in Davis et al. 1989; Venkatesh and Davis 1996). 122 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage Table 1. Summary of Hypotheses Relationship Hypothesis Short-term Effects H1 U-BI Men > Women H2a EOU-BI Women > Men H2b EOU-U Men > Women H3 SN-BI Women > Men Long-term Effects H4 U-BI Men > Women H5a EOU-BI Women > Men H5b EOU-U Men > Women H6 SN-BI Non-significant Given the authors’ prior agreement with the field or its objectives. User reactions to the technology sites, all members of the relevant departments were gathered across three points in time: where the new system was being introduced immediately after the initial training (t ), after one were participants in this research study. A 77% month of experience (t ), and after three months of response rate (342 usable responses including experience (t ). Actual usage behavior (USE) was 156 women and 186 men) was achieved across measured over the five-month period from the all three points of measurement. The responses time of initial introduction of the technology. For for any one individual were dropped if responses purposes of this research, t represented the from that individual were not received for all three measurement point to study short-term effects periods. On average, participants had an (i.e., initial user reactions), and t and t repre- 2 3 average of 5.5 years of prior experience using sented measurements to study long-term effects computers, with a range from six months to 16 (i.e., situations of significant direct experience with years. As expected, based on a pre-study the technology). U, EOU, and SN measured in a questionnaire, it was found that none of the specific time period (e.g., t ) were used to predict participants had any prior knowledge about the intention measured in the same time period. system being introduced. Intention measured in a given time period was used to predict subsequent usage behavior. Figure 2 presents a summary of the design and points of measurement of this research. Procedure and Measurement Validated items were used to measure perceived User reactions and usage behavior were mea- usefulness, perceived ease of use, subjective sured over a period of five months. Participants norm, and behavioral intention (Davis 1989; Davis in each organization participated in a one-day et al. 1989; Mathieson 1991; Taylor and Todd training program on the system. The training 1995a, 1995b). Actual usage behavior (USE), included two hours of lecture, followed by two operationalized as the frequency of use (number hours of lecture combined with hands-on use, of user queries for information), was gathered and two hours of independent interaction with the from system logs. Consistent with prior research system (with consultants being available for in sociology and organization behavior, we mea- help). Between 20 and 25 participants were sured demographic variables of interest: gender, included in each session, with multiple sessions of training being conducted in each organization. income, education, and organizational position. Neither the lecturers nor the training assistants Appendix A presents a list of the items used in this (software consultants) knew about the research research. MIS Quarterly Vol. 24 No. 1/March 2000 123 Venkatesh & Morris/Gender in Technology Acceptance and Usage Short-term effects Long-term effects Initial Follow-up Follow-up D ata gathering m easures gathered m easures gathered m easures gathered com pleted use use use training t t t 1 2 3 (post training) (one m onth) (three m onths) (fiv e m onths) Figure 2. Research Design and Timing of Measurement Results constructs pertaining to TAM (e.g., Adams et al. 1992) and subjective norm (e.g., Mathieson 1991) have been extensively tested and validated in Measurement Model Estimation prior research. Once the measurement models were found to be Partial Least Squares (PLS) was used to analyze acceptable, it was important to ascertain if the the data. PLS is an extremely powerful structural structural models were comparable across equation modeling (SEM) technique that has been organizations. This was considered particularly used extensively in information systems research important if the data were to be pooled across (see Chin et al. 1996; Compeau and Higgins organizations. To examine this issue, the data 1995a, 1995b; Sambamurthy and Chin 1994). were pooled across organizations at each of the The software package used to perform the analy- three points of measurement and dummy vari- sis was PLS Graph, Version 2.91.03.04. ables were introduced to distinguish data from the different organizations. The coding scheme used The measurement model was assessed sepa- four dummy variables (DUMMY1, DUMMY2, rately for each of the five organizations at each of DUMMY3, and DUMMY4) that were coded as the three different points of measurement, thus follows: 0,0,0,0 to represent organization #1; resulting in 15 examinations. All constructs in all 0,0,0,1 to represent organization #2; 0,0,1,0 to models satisfied the criteria of reliability (ICR > represent organization #3; 0,1,0,0 to represent .80) and discriminant validity (shared variance organization number #4; 1,0,0,0 to represent across items measuring a construct was higher organization #5. In this case, each of the dummy than correlations across constructs), thus re- variables employed represented their own latent quiring no changes to the constructs. The basic variable with one indicator variable with a factor structure of the measurement model analyses was loading of 1.00. In addition to the main effects, consistent across all 15 estimations. This pattern interaction terms incorporating the dummy of high reliability and validity was consistent with variables to represent organizations (e.g., U X our expectations given that the scales for the 124 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage DUMMY1 X DUMMY2 X DUMMY3 X DUMMY4) prior experience with computers was also were introduced in models estimated for the entire included. The direct effect of each of these sample, women only, and men only at each of the variables on model relationships was examined three points of measurement. If any of the (e.g., effect of INCOME on U-BI) as well as interaction terms were significant, it would indicate interactive effects with gender (e.g., U-BI as differences in the corresponding structural path moderated by both INCOME and GENDER). All coefficients across different organizations. In the tests for confounding effects were non-significant, current data set, none of the interaction terms thus demonstrating that income, occupation, were significant, suggesting that the results from educational level, and prior experience did not each of the five organizations were statistically confound gender differences. equivalent. Armed with the high degree of consistency in the measurement and structural model analyses across organizations and Hypotheses Testing consistent with prior research (e.g., Compeau and Higgins 1995b; Davis et al. 1989; Venkatesh and Table 1 summarizes the hypotheses being tested. Davis 1996), we pooled the data across For the purpose of this research, we expect that organizations to increase power and facilitate the short-term vs. long-term differences will help brevity of results reporting. us glean an understanding of the influence of experience in this context. To that end, using the The results of the measurement model estimation different points of measurement as a proxy for based on the data pooled across organizational user experience with the system is consistent with sites at t are summarized in Appendix B (B1 prior research (e.g., Davis et al. 1989; Venkatesh reports the factor structure and B2 reports the and Davis 1996). reliability and discriminant validity coefficients). The pattern of measurement model results was At each of the three points of measurement, the consistent at the other two points of measurement structural model was tested with the data from the as well. Table 2 presents the descriptive statistics entire sample (i.e., women and men pooled (means and standard deviations) of the different together) and each of the subsamples (i.e., variables, categorized by gender, at each point of women taken separately and men taken sepa- measurement. With the exception of U at t and rately). Table 3 presents the path coefficients for SN at t (SN was a non-significant determinant of each of the subsamples so that the reader may intention at t —see hypothesis testing, discussed clearly see the magnitude of any differences—and later), the mean values between women and men thus the practical significance—between men and were statistically significantly different (p < .05) at women across each of the constructs. Following all three points of measurement. the model tests, we conducted a test of the dif- ferences in path coefficients between the two subsamples; also, we conducted a test of the Pretesting Checks for Potential differences in path coefficients between each of Confounds the subsamples and the entire sample. In prior organizational behavior research, a After initial exposure, compared to women, men number of demographic variables have been placed a greater emphasis on U in determining BI, shown to potentially confound observed gender as hypothesized (H1). On the other hand, women differences. Income, occupation, and educational weighted EOU more strongly in determining BI levels are considered to be the most important than men did at t , consistent with H2a. In fact, confounds (see Lefkowitz 1994 for a discussion). EOU was not a significant determinant of BI for In addition, prior experience with computers is a men, possibly due to variance suppression in variable that could possibly confound gender EOU (SD = 0.6). Contrary to H2b, there were no differences in technology perceptions and usage. gender differences in the role of EOU in Therefore, in addition to the confounding variables determining U. Finally, in the short term, SN was identified in organizational behavior research, a significant factor influencing BI for women after MIS Quarterly Vol. 24 No. 1/March 2000 125 Venkatesh & Morris/Gender in Technology Acceptance and Usage Table 2. Descriptive Statistics by Gender Significance of Women Men Difference Between MSD M SD Women and Men Post Training U 4.5 1.1 5.0 1.0 ns EOU 4.2 0.8 5.3 0.6 * SN 4.1 0.8 5.0 0.8 * BI 3.8 1.0 5.1 1.1 ** After one month U 4.2 1.2 5.1 0.8 * EOU 3.9 1.1 5.7 0.9 *** SN 4.4 1.0 5.1 0.7 * BI 3.6 0.8 4.9 1.2 ** USE 4.7 1.1 8.8 2.0 ** After three months U 4.1 1.0 5.2 0.7 * EOU 3.7 1.0 5.7 0.8 *** SN 4.1 0.9 4.1 0.8 ns BI 3.7 1.1 5.0 0.8 ** USE 5.9 1.4 11.2 2.8 *** USE 6.2 1.3 10.1 3.2 *** Notes: 1. Use refers to the average weekly usage between measurement 1 (post training) and measurement 2 (after one month), Use refers to the average weekly usage between measurement 2 (after one month) and measurement 3 (after three months), and Use refers to the average weekly usage between measurement 3 (after three months) and measurement 4 (after five months). 2. Weekly usage is reported so as to allow a direct comparison of usage across time periods (t - t , t - 1 2 2 t , and t - t ) since the time lapsed in each interval is different. 3 3 4 3. The significance of difference column reports the results corresponding to an independent samples difference of means test. ns: non-significant; * p < .05; ** p < .01; *** p < .001 126 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage Table 3. Gender Differences in the Salience of Perceived Usefulness, Perceived Ease of Use, and Subjective Norm in Determining Behavioral Intention Entire Diff Diff Diff Sample Women Men Sample Sample Women vs. vs. vs. 2 2 2 R A R A R A Women Men Men Time 1 .41 .42 .40 U-BI .47*** .30*** .61*** *** *** *** EOU-BI .20** .33*** .10 * * ** SN-BI .12* .33*** .08 ** * ** EOU-U .18** .20** .18** ns ns ns Time 2 .40 .40 .39 U-BI .49*** .32*** .62*** *** *** *** EOU-BI .18* .31*** .01 * * ** SN-BI .14* .33*** .04 ** * ** EOU-U .18** .21** .19** ns ns ns Time 3 .41 .42 .40 U-BI .51*** .36*** .62*** ** *** *** EOU-BI .21** .36*** .05 * ** *** SN-BI .04 .10 .09 ns ns ns EOU-U .20** .20** .20** ns ns ns Notes: 1. The three difference columns present the significance of difference in path coefficients between the entire sample and subsample of women, the entire sample and subsample of men, and the subsamples of women and men respectively. Specifically, the significance of difference was calculated using the procedure described in Cohen and Cohen (1988, pp. 55-56). 2. The R reported corresponds to the structural equations BI = U + EOU + SN. The EOU-U path coefficient is reported from the structural equation U = EOU. The R corresponding to the EOU-U path in each case is the square of the coefficient reported. ns: non-signifcant; * p < .05; ** p < .01; *** p < .001 MIS Quarterly Vol. 24 No. 1/March 2000 127 Venkatesh & Morris/Gender in Technology Acceptance and Usage initial training; however, SN did not play a al. (1988), who found an intention-behavior significant role in determining BI among men, correlation of about 0.50 based on a meta- providing support for H3. analysis of 87 studies and recent technology acceptance research (Venkatesh and Speier Over the long term, men were more strongly 1999). influenced by U in determining BI, compared to women, as hypothesized (H4). Similarly, women continued to weight EOU as a direct determinant Discussion of BI more strongly than men, providing support for H5a. Consistent with the results in the short This research has addressed the question: “Are term and contrary to H5b, there were no men and women different with respect to techno- differences in the EOU-U relationship between logy adoption?” Rather than examining mean men and women. While SN did not influence men differences between women and men, this at t and t (partially supporting H6), women were 2 3 research focused on a longitudinal examination of still influenced by subjective norm after one month gender differences in the relationships among of sustained technology use (t ), contrary to H6. theoretically grounded determinants of technology The increased salience of subjective norm at t acceptance and usage. The focus on the relative and t is particularly interesting given the some- influence of different determinants (beta dif- what lower level of perceived normative pressure ferences) demonstrates how women and men among women compared to men (see Table 2). differ in their decision making processes regarding However, the salience of SN for women became technology acceptance and use. Several impor- non-significant at t , as predicted. The support for tant and interesting findings, both over the short- the null hypothesis in that subjective norm was not and long-term, regarding the roles of perceived a determinant at t calls for a power test to usefulness, perceived ease of use, and subjective understand the potential for type II error (Cohen norm emerged from this work. 1988). We found the power to be just under 0.85 for small effects and over 0.90 for medium effects, The current research revealed that men consider thus largely alleviating concerns about type II perceived usefulness to a greater extent than error. Table 4 summarizes the results of the women in making their decisions regarding the hypotheses testing. use of a new technology, both in the short- and long-term. On the other hand, perceived ease of To enhance the nomological validity of the use was more salient to women compared with findings, we examined how usage behavior fit with men both after initial training and over time with the proposed extensions to TAM. Usage data increasing experience with the system. In fact, gathered in the time period from t to t was used 1 2 perceived ease of use was not a salient factor to as the dependent variable in the structural model men at any point in time. Interestingly, men’s corresponding to t ; similarly, usage data gathered assessment of ease of use of the system went up from t to t was used as the dependent variable in 2 3 somewhat with time/experience and women’s the structural model corresponding to t , and 2 assessment went down. This adds further evi- usage data gathered for two months after t was dence to the differential salience observed used to test the model corresponding to t . The 3 because men perceive the system to be easier to direct path coefficients between the determinant use with increasing experience, thus resulting in beliefs and usage behavior were examined. The perceptions of ease of use receding into the direct paths from U, EOU, and SN to usage background and being a non-significant factor in behavior were found to be non-significant in all determining their intention to use the system. In cases (women and men at all points of mea- contrast, the declining perceptions of ease of use surement), thus indicating that the effects of U, of the system observed in women appear to make EOU, and SN on usage behavior were fully system ease of use more of an issue to them, thus mediated by behavioral intention. The intention- to some extent accounting for the increased behavior path coefficient was found to be between salience of ease of use for women relative to other 0.49 and 0.56. This consistent with Sheppard et usage determinants. 128 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage Table 4. Summary of Results Relationship Hypothesis Remarks Short-term Effects H1 U-BI Men > Women Supported H2a EOU-BI Women > Men Supported H2b EOU-U Men > Women Not supported H3 SN-BI Women > Men Supported Long-term Effects H4 U-BI Men > Women Supported H5a EOU-BI Women > Men Supported H5b EOU-U Men > Women Not supported H6 SN-BI Non-significant Partially supported (significant for women at t ) For subjective norm, the contrasts were equally Based on our results, several important inferences striking. Subjective norm did not influence men’s can be made. First, given the findings, one could decisions at any point in time. In contrast, women argue that men are more driven by instrumental did consider normative influences at the initial factors (i.e., perceived usefulness) while women stage of technology introduction and after one are more motivated by process (perceived ease of month of experience. After three months of use) and social (subjective norm) factors. How- experience, women no longer placed significant ever, perhaps a more qualitative interpretation emphasis on subjective norm. This outcome was would suggest that men are more focused in their contrary to our expectation that subjective norm decision making regarding new technologies, would not be significant with increasing while women are more balanced in their decision- experience (i.e., during measurement after one making process. In other words, while men only and three months of use) due to internalization of consider productivity-related factors, women con- normative influences. One possible explanation sider inputs from a number of sources including for this outcome is that one month (t ) was not productivity assessments when making techno- enough time to gain direct experience that leads logy adoption and usage decisions. This notion is to cementing of one’s own views regarding the new system. Women may still have been supported by the fact that all three determinants receiving and considering input from peers/ (U, EOU, and SN) together explain nearly identical superiors and had not fully internalized others’ variance in initial intention for women as perceived views. However, it appears that three months usefulness (U) alone explains in initial intention for (i.e., t ) was long enough for internalization to take 3 men. This basic pattern held true in explaining place, rendering subjective norm non-significant. sustained usage of technology as well. Further- Usage statistics (see Table 2) indicated that it is more, these gender differences were robust to the possible that this outcome occurred because the most important potential confounds of gender frequency of usage by women was about half the studies in the organizational behavior research use by men. Interestingly, although women, in and technology research, thus providing com- contrast to men, considered normative influences pelling evidence for the notion that gender plays a in their decision making process, the perceptions vital role in shaping initial and sustained techno- of normative pressure among women were logy adoption decisions by today’s knowledge actually lower than the perceived pressure among workers. men. MIS Quarterly Vol. 24 No. 1/March 2000 129 Venkatesh & Morris/Gender in Technology Acceptance and Usage Contributions and Implications fluences. As discussed earlier, Minton and Schneider (1980) and Roberts (1991) suggest two The current research presents important con- potentially competing causal mechanisms. tributions and implications for research and Although both lines of argument suggest similar practice. TAM has been replicated and applied in outcomes, the information processing models pro- a wide variety of settings for nearly a decade. posed are different. It is important to understand However, extensions to the model have been the circumstances in which different mechanisms limited. Specifically, research has not yet investi- are operational in order to facilitate the design of gated the “conditions and mechanisms governing appropriate organizational interventions for the impact of social influences on usage behavior” increased buy-in for technologies being intro- called for by Davis et al. (1989, p. 999). Thus, the duced. More broadly, it is important to understand proposed extensions to TAM—the integration of the cognitive mechanisms underlying the forma- subjective norm, examination of gender dif- tion and change of perceived usefulness and ferences in the role of the original TAM constructs, perceived ease of use in general (see Davis et al. and the related role of experience—represent 1992; Venkatesh and Davis 1996), and among important theoretical advances in technology women and men separately. acceptance and usage. The current research integrates subjective norm into TAM and delin- Much prior research on TAM has presented a eates when subjective norm will play a role from cross-sectional snapshot (e.g., Mathieson 1991), the perspective of target user category (i.e., or has used student subjects in a longitudinal women) and timing (i.e., short-term rather than study (e.g., Venkatesh and Davis 1996). Thus, long-term). Further, identifying boundary condi- one important strength of this research is the tions (i.e., moderation by gender) associated with longitudinal nature (five months) of the study the role played by the original TAM constructs of combined with the real organizational contexts perceived usefulness and perceived ease of use (five different organizations) to study user helps us refine, sharpen, and, quite possibly, reactions and usage behavior. In a real-world better apply TAM to the study of user acceptance setting, this research presented the opportunity to and usage in the workplace. The robustness of study user reactions to a new technology as users the findings over a five-month period in a real- progressed from novices on the new system to world setting provides strong evidence supporting experienced users. The findings, therefore, help the proposed extensions and boundary conditions. us better understand gender differences in The basic TAM hypothesis that the effect of technology acceptance, adoption, and usage, thus external variables (e.g., gender) will be completely providing valuable insights into implementation mediated, with no moderating effects, was not and diffusion of new technologies in organizational supported. Such a pattern is consistent with settings. The current work combined with our psychology research (e.g., Tashakkori and other recent work (Venkatesh et al. forthcoming), Thompson 1991). This calls for research into which presents a longitudinal analysis, provides a other situations and circumstances when there is more complete picture of gender and technology partial mediation of external variables by TAM adoption/usage. Unfortunately, the role of age constructs, and the need to identify other potential could not be studied due to restrictions imposed moderators and boundary conditions of TAM. by the participating organizations. However, in other work, we have studied the role of age but The importance of subjective norm in determining not gender, once again due to practical technology adoption decisions among women constraints (Morris and Venkatesh forthcoming). merits further attention by researchers and Future research should examine the role of practitioners alike. Peer pressure and superiors’ gender and age in the context of a single research influence have been shown to be determinants of study. subjective norm in technology adoption contexts. Future research should focus on clarifying the There are also important practical implications for underlying cognitive mechanisms for the greater these findings. Organizations today invest over importance placed by women on normative in- $20 billion in technology training programs 130 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage (Industry Report 1996). Training represents the research should measure expressiveness, aware- key method for successful knowledge transfer to ness of others’ feelings, and motivation to comply users, implementation, and diffusion of new to examine the underlying psychological dimen- technologies, and is the most popular mechanism sions captured via gender. This would be useful used to smooth the transition to new technology in for several reasons. First, men and women are the workplace. However, if such training pro- not at bipolar extremes on these dimensions. grams are to be effective in helping organizations Thus, they might vary based on degrees of overcome barriers to adoption, the current femininity or masculinity (Bem 1981). Further- research suggests that trainers are faced with a more, TAM is a psychological model. While the dilemma: Do they emphasize productivity benefits, consideration of gender as a biological construct or do they emphasize process/usability issues and in this research is consistent with previous social factors? Trainers should be careful not to conceptualizations of the construct, it adds a layer treat this issue as a “zero sum game” (i.e., of abstraction to TAM that might be alleviated by emphasizing one factor at the expense of a psychological examination of gender or its another). Rather, they may wish to emphasize underlying dimensions in future research. usefulness issues for men, while offering women a more balanced analysis that includes produc- Another measurement limitation was the opera- tivity aspects, process issues, and testimonials tionalization of the prior computer experience from peers or superiors. These recommendations construct in this study. The construct was mea- also have implications for marketing professionals sured by the number of years of experience the who may find these findings useful in designing user had with computers in general. Because advertising campaigns designed to appeal to a none of the participants had any prior experience specific target group within the population. Again, with the target system, we believe the experience by targeting outcome expectations vs. process measure used in this study was reasonable. expectations and/or social factors, one may pin- Future research might use a finer grain of detail in point important issues related to technology its conceptualization of experience. For example, adoption for men and women, respectively. The two years using solely word processing is much overall pattern of gender differences also presents different from two years of programming organizations with important information in terms experience. Future research might also target of designing organizational and managerial self-efficacy (Compeau and Higgins 1995a, interventions that can foster acceptance and use 1995b) or domain-specific experience as alter- of new technologies both in the short- and the native measures to employ. Another limitation in long-term. the current work is the measurement of usage as frequency of use. While there are precedents to such a measurement of usage (e.g., Davis et al. 1989), future research should employ duration of Limitations and Future Research use and/or other measures that more completely Directions capture the intensity of usage. One potential limitation of this research surrounds A number of other measurement issues with the measure of gender employed. The dichoto- respect to the demographic variables employed in mous measurement is consistent with the treat- this study offer important avenues for extensions ment of gender as “biological sex.” As noted in of this work. Different categorizations of the occu- the literature review, gender may also be pational variable (for example, into technical and conceptualized as a psychological construct (e.g., non-technical) may be valuable. Educational level Bem 1981). If so, gender (as operationalized in could measure domain-specific knowledge (e.g., this study) may be a surrogate for other computer aptitude tests) or more generalized psychological constructs. For example, our measures of intelligence (e.g., IQ tests) to extend research suggests that subjective norm is more the educational level as was measured in this important for women because, as a group, they research. Income level could also be operationa- are more expressive, more aware of others’ feelings, and more compliant than men. Future lized as household income given the prevalence MIS Quarterly Vol. 24 No. 1/March 2000 131 Venkatesh & Morris/Gender in Technology Acceptance and Usage of dual incomes today. Such a conceptualization References may more accurately reflect socialization patterns Adams, D. A., Nelson, R. R., and Todd, P. A. and schematic processing related to socio- “Perceived Usefulness, Ease of Use and Usage economic status. of Information Technology: A Replication,” MIS Quarterly (16:2), 1992, pp. 227-250. Ajzen, I. “From Intentions to Actions: A Theory of Conclusions Planned Behavior,” in Action Control: From Cog- nition to Behavior, J. Kuhl and J. Beckmann (eds.), Springer Verlag, New York, 1985, pp. 11- While TAM is a parsimonious and powerful model for understanding technology acceptance, social Ajzen, I. “The Theory of Planned Behavior,” influence is a notable omission from the model. Organizational Behavior and Human Decision This research addresses this issue by Processes (50), 1991, pp. 179-211. investigating gender as a potential key to Ajzen, I., and Fishbein, M. Understanding Attitudes understanding the role of social influence in initial and Predicting Social Behavior, Prentice-Hall, technology adoption decisions and sustained Englewood Cliffs, NJ, 1980. usage of new technologies. In addition to gender Bandura, A. “Self-Efficacy: Toward a Unifying differences in the role of social influence, the Theory of Behavioral Change,” Psychological current research also reveals gender differences Review (84:2), 1977, pp. 191-215. in the role of perceived usefulness and perceived Bandura, A. “Self-Efficacy Mechanism in Human ease of use as determinants of technology Agency,” American Psychologist (37:2), 1982, acceptance and usage behavior. Importantly, the pp. 122-147. gender differences reported here were robust to Bandura, A. Social Foundations of Thought and the key confounds identified in prior organizational Action: A Social Cognitive Theory, Prentice-Hall, Englewood Cliffs, NJ, 1986. behavior literature and information technology Barnett, R. C., and Marshall, N. L. “The Rela- research, thus further supporting the proposed tionship Between Women’s Work and Family extension to TAM. While men still represent a Roles and Their Subjective Well-Being and majority of the work force, the number of women Psychological Distress,” in Women, Work and at all levels of the organizational hierarchy Health: Stress and Opportunities, M. Franken- continues to rise. Therefore, technology accep- haeuser, V. Lundberg, and M. A. Chesney (eds.), tance theories and models that overlook gender Plenum, New York, 1991, pp. 111-136. as an important factor may overestimate the Becker, B. J. “Influence Again: An Examination of influence of productivity-oriented factors while Reviews and Studies of Gender Differences in simultaneously underestimating the importance of Social Influence,” in The Psychology of Gender: ease of use perceptions and social influences. Advances Through Meta-Analysis, J. S. Hyde and M. C. Linn (eds.), Johns Hopkins University Press, Baltimore, MD, 1986, pp. 178-209. Bem, S. L. “The BSRI and Gender Schema Acknowledgements Theory: A Reply to Spence and Helmreich,” Psychological Review (88), 1981, pp. 369-371. We are indebted to the senior editor, Professor Bem, D. J., and Allen, A. “On Predicting Some of Ron Weber, the anonymous associate editor, and the People Some of the Time: The Search for three reviewers whose insightful comments and Cross-Situational Consistencies in Behavior,” suggestions significantly enhanced the quality of Psychological Review (81), 1974, pp. 506-520. this paper. We would like to express our sincerest Bergeron, F., Rivard, S., and De Serre, L. “Investi- thanks to Tracy Ann Sykes (University of Mary- gating the Support Role of the Information land) for her editorial work that helped improve the Center,” MIS Quarterly (14:3), 1990, pp. 247- readability of this paper. 259. 132 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage Bettman, J. R., and Sujan, M. “Effects of Framing Davis, F. D. “Perceived Usefulness, Perceived on Evaluations of Comparable and Noncom- Ease of Use, and User Acceptance of Infor- parable Alternatives by Expert and Novice mation Technology,” MIS Quarterly (13:3), 1989, Consumers,” Journal of Consumer Research pp. 319-339. (14:2), 1987, pp. 141-154. Davis, F. D. “User Acceptance of Information Bozionelos, N. “Psychology of Computer Use: Technology: System Characteristics, User Per- XXXIX. Prevalence of Computer Anxiety in British ceptions and Behavioral Impacts,” International Managers and Professionals,” Psychological Journal of Man-Machine Studies (38:3), 1993, pp. Reports (78), 1996, pp. 995-1002. 475-487. Carlson, R. “Sex Differences in Ego Functioning: Davis, F. D., Bagozzi, R. P., and Warshaw, P. R. Exploratory Studies of Agency and Communion,” “User Acceptance of Computer Technology: A Journal of Consulting and Clinical Psychology Comparison of Two Theoretical Models,” (37), 1971, pp. 267-277. Management Science (35), 1989, pp. 982-1002. Chan, D. K. S., and Fishbein, M. “Determinants of Davis, F. D., Bagozzi, R. P., and Warshaw, P. R. College Women’s Intention to Tell Their Partners “Extrinsic and Intrinsic Motivation to Use Com- to Use Condoms,” Journal of Applied Social puters in the Workplace,” Journal of Applied Psychology (23), 1993, pp. 1455-1470. Social Psychology (22:14), 1992, pp.1111-1132. Chin, W. W., Marcolin, B. L., and Newsted, P. R. Davis, F. D., and Venkatesh, V. “A Critical Assess- “A Partial Least Squares Latent Variable ment of Potential Measurement Biases in the Modeling Approach for Measuring Interaction Technology Acceptance Model: Three Experi- Effects: Results from a Monte Carlo Simulation ments,” International Journal of Human Study and Voice Mail Emotion/Adoption Study,” Computer Studies (45), 1996, pp. 19-45. in Proceedings of the Seventeenth International Doll, J., and Ajzen, I. “Accessibility and Stability of Conference on Information Systems, J. I. Predictors in the Theory of Planned Behavior,” DeGross, S. Jarvenpaa, and A. Srinivasan (eds.), Journal of Personality and Social Psychology Cleveland, OH, 1996, pp. 21-41. (63), 1992, pp. 754-765. Chin, W. W., and Todd, P. A. “On the Use, Useful- Eagly, A. H. “Sex Differences in Influenceability,” ness, and Ease of Use of Structural Equation Psychological Bulletin (85), 1978, pp. 86-116. Modeling in MIS Research: A Note of Caution,” Eagly, A. H., and Carli, L. L. “Sex of Researchers MIS Quarterly (19), 1995, pp. 237-246. and Sex-Typed Communications as Deter- Cohen, J. Statistical Power Analysis for the minants of Sex Differences in Influenceability: A Behavioral Sciences, Lawrence Erlbaum Asso- Meta-Analysis of Social Influence Studies,” ciates, Hillsdale, NJ, 1988. Psychological Bulletin (90), 1981, pp. 1-20. Cohen, J., and Cohen, P. Applied Multiple Regres- Fazio, R. H., and Zanna, M. “Attitudinal Qualities sion/Correlation Analysis for the Behavioral Relating to the Strength of the Attitude- nd Sciences (2 ed.), Lawrence Erlbaum Asso- Perception and Attitude-Behavior Relationship,” ciates, Hillsdale, NJ, 1988. Journal of Experimental Social Psychology (14), Compeau, D. R., and Higgins, C. A. “Application of 1978a, pp. 398-408. Social Cognitive Theory to Training for Computer Fazio, R. H., and Zanna, M. “On the Predictive Skills,” Information Systems Research (6:2), Validity of Attitudes: The Roles of Direct Exper- 1995a, pp. 118-143. ience and Confidence,” Journal of Personality Compeau, D. R., and Higgins, C. A. “Computer (46), 1978b, pp. 228-243. Self-Efficacy: Development of a Measure and Fazio, R. H., and Zanna, M. “Direct Experience and Initial Test,” MIS Quarterly (19:2), 1995b, pp. Attitude-Behavior Consistency,” in Advances in 189-211. Experimental Social Psychology (14) L. Berko- Crawford, M., Chaffin, R., and Fitton, L. “Cognition witz (ed.), Academic Press, San Diego, CA, in Social Context,” Special Issue, Psychological 1981, pp. 161-202. and Psychobiological Perspectives on Sex Felter, M. “Sex Differences on the California State- Differences in Cognition: I. Theory and Research, Sex Learning and Individual Differences (7:4), 1995, wide Assessment of Computer Literacy,” pp. 341-362. Roles (13), 1985, pp. 181-192. MIS Quarterly Vol. 24 No. 1/March 2000 133 Venkatesh & Morris/Gender in Technology Acceptance and Usage Fishbein, M., and Ajzen, I. Belief, Attitude, Intention Overwhelming Technological Change, Addison- and Behavior: An Introduction to Theory and Wesley, Reading, MA, 1996. Research, Addison-Wesley, Reading, MA, 1975. Kagan, J. “Acquisition and Significance of Sex- Fishbein, M., and Stasson, M. “The Role of Typing and Sex Role Identity,” in Review of Child Desires, Self-Predictions, and Perceived Control Development Research (1), M. L. Hoffman and L. in the Prediction of Training Session Attendance,” W. Hoffman (eds.), Russell Sage Foundation, Journal of Applied Social Psychology (20), 1990, New York, 1964, pp. 137-167. pp. 173-198. Kohlberg, L. A. “A Cognitive-Developmental Analy- Fulk, J., Steinfield, C. W., Schmitz, J., and Power, sis of Children's Sex-Role Concepts and Atti- J. G. “A Social Information Processing Model of tudes,” in The Development of Sex Differences, Media Use in Organizations,” Communications E. E. Maccoby (ed.), Stanford University Press, Research (14), 1987, pp. 529-552. Stanford, CA, 1966, pp. 82-173. Garai, J. E., and Scheinfeld, A. “Sex Differences in Landau, S. B., and Leventhal, G. A. “A Simulation Mental and Behavioral Traits,” Genetic Psycho- Study of Administrators' Behavior Toward Em- logy Monographs (77), 1968, pp. 169-299. ployees Who Receive Job Offers,” Journal of Gefen, D., and Straub, D. W. “Gender Differences Applied Social Psychology (6), 1976, pp. 291- in the Perception and Use of E-mail: An Exten- 306. sion to the Technology Acceptance Model,” MIS Landauer, T. K. The Trouble with Computers: Use- Quarterly (21:4), 1997, pp. 389-400. fulness, Usability, and Productivity, MIT Press, Gill, S., Stockard, J., Johnson, M., and William, S. Cambridge, MA, 1995. “Measuring Gender Differences: The Expressive Lapan, R. T., McGrath, E., and Kaplan, D. “Factor Dimension and Critique of Androgyny Scales,” Structure of the Basic Interest Scales by Gender Sex Roles (17), 1987, pp. 375-400. Across Time,” Journal of Counseling Psychology Hartwick, J., and Barki, H. “Explaining the Role of (37:2), 1990, pp.216-222. User Participation in Information System Use,” Lee, A. “Electronic Mail as a Medium for Rich Management Science (40), 1994, pp. 440-465. Communication: An Empirical Investigation Using Hoffman, L. W. “Early Childhood Experiences and Hermeneutic Interpretation,” MIS Quarterly Women's Achievement Motives,” Journal of (18:2), 1994, pp. 143-157. Social Issues (28:2), 1972, pp. 129-155. Lefkowitz, J. “Sex-Related Differences in Job Attit- Hoffman, L. W. “Changes in Family Roles, Sociali- udes and Dispositional Variables: Now You See zation, and Sex Differences,” American Psycho- Them....”, Academy of Management Journal logist (32), 1977, pp. 644-657. (37:2), 1994, pp. 323-349. Hofstede, G. Culture’s Consequences: Interna- Lueptow, L. B., Garovich, L., and Lueptow, M. B. tional Differences in Work Related Values, Sage, “The Persistence of Gender Stereotypes in the London, 1980. Face of Changing Sex Roles: Evidence Contrary Hunt, N. P., and Bohlin, R. M. “Teacher Education to the Sociocultural Model,” Ethnology and and Students’ Attitudes Toward Using Com- Sociobiology (16), 1995, pp. 509-530. puters,” Journal of Research on Computing in Lundeberg, M. A., Fox, P. W., and Puncochar, J. (25), 1993, pp. 487-497. Education “Highly Confident But Wrong: Gender Differences Igbaria, M., and Chakrabarti, A. “Computer Anxiety and Similarities in Confidence Judgments,” and Attitudes Toward Microcomputer Use,” Be- Journal of Educational Psychology (86), 1994, haviour and Information Technology (9:3), 1990, pp. 114-121. pp. 229-241. Maccoby, E. E., and Jacklin, C. N. The Psychology Igbaria, M., Zinatelli, N., Cragg, P., and Cavaye, A. of Sex Differences, Stanford University Press, L. M. “Personal Computing Acceptance Factors Stanford, CA, 1974. in Small Firms: A Structural Equation Model,” Markus, M. L. “Toward a ‘Critical Mass’ Theory of MIS Quarterly (21:3), 1997, pp. 279-305. Interactive Media,” in Organizations and Com- Industry Report. Training (33:10) 1996, pp.37-79. Johansen, R., and Swigart, R. Upsizing the Indi- munication Technology, J. Fulk and C. W. vidual in the Downsized Organization: Managing Steinfield (eds.), Sage, Newbury Park, CA, 1990, in the Wake of Reengineering, Globalization, and pp. 491-511. 134 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage Mathieson, K. “Predicting User Intentions: Com- Parsons, J. E., and Bales, R. F. Family: Sociali- zation and Interaction Process, Free Press, paring the Technology Acceptance Model with Glencoe, IL, 1955. the Theory of Planned Behavior,” Information Poole, M. S., and DeSanctis, G. “Understanding Systems Research (2), 1991, pp. 173-191. the Use of Group Decisions Support Systems: Mervis, C. B., and Rosch, E., “Categorization of The Theory of Adaptive Structuration,” in Organi- Natural Objects,” Annual Review of Psychology zations and Communication Technology, J. Fulk, (32), 1980, pp. 89-115. and C. Steinfield (eds.), Sage, Newbury Park, Miller, J. B. Toward a New Psychology of Women, CA, 1990, pp.173-191. Beacon Press, Boston, 1986. Regan, D. T., and Fazio, R. H. “On the Consistency Minton, H. L., and Schneider, F. W. Differential Between Attitudes and Behavior: Look to the Psychology, Waveland Press, Prospect Heights, Method of Attitude Formation,” Journal of Experi- IL, 1980. mental Social Psychology (13), 1977, pp. 38-45. Minton, C., Kagan, J., and Levine, J. A. “Maternal Reinecke, J., Schmidt, P., and Ajzen, I. Control and Obedience in the Two-year-old,” “Application of the Theory of Planned Behavior to Child Development (42), 1971, pp. 1873-1894. Adolescents’ Condom Use: A Panel Study,” Moore, G. A. Crossing the Chasm: Marketing and Journal of Applied Social Psychology (26), 1996, Selling High-Tech Products to Mainstream pp. 749-772. Roberts, T. “Gender and the Influence of Customers, Harper-Collins, New York, 1991. Evaluations in Self-Assessments in Achievement Morris, M., and Dillon, A. “How User Perceptions Settings,” Psychological Bulletin (109), 1991, pp. Influence Software Use,” IEEE Software (14:4), 297-308. 1997, pp. 58-65. Rosen, L. D., and Maguire, P. D. “Myths and Morris, M. G., and Venkatesh, V. “Age Differences Realities in Computerphobia: A Meta-Analysis,” in Technology Adoption Decisions: Implications Anxiety Research (3), 1990, pp. 175-191. for a Changing Work Force,” Personnel Psycho- Rosenkrantz, P., Vogel, P., Bee, H., Broverman, I., logy, forthcoming. and Broverman, D. M. “Sex-Role Stereotypes Morrow, P. C., Presll, E. R., and McElroy, J. C. and Self-Concepts in College Students,” Journal “Attitudinal and Behavioral Correlates of Com- of Consulting and Clinical Psychology (32), 1968, puter Anxiety,” Psychological Reports (59), 1986, pp. 287-295. pp. 1199-1204. Sambamurthy, V., and Chin, W. W. “The Effects of Ngwenyama O. J., and Lee A. S. “Communication Group Attitudes Toward Alternative GDSS Richness in Electronic Mail: Critical Social Designs on the Decision-Making Performance of Theory and the Contextuality of Meaning,” MIS Computer-Supported Groups,” Decision Sciences (25:2), 1994, pp. 215-241. Quarterly (21:2), 1997, pp. 145-168 Segars, A. H., and Grover, V. “Re-examining Per- Nisbett, R. E., and Ross, L. Human Inference: ceived Ease of Use and Usefulness: A Strategies and Shortcomings of Social Judge- Confirmatory Factor Analysis,” MIS Quarterly ment, Prentice-Hall, Englewood Cliffs, NJ, 1980. (18:4), 1993, pp. 517-525. Norman, D. A. Things That Make Us Smart: Sheppard, B. H., Hartwick, J., and Warshaw, P. R. Defending Human Attributes in the Age of the “The Theory of Reasoned Action: A Meta-Analy- Machine, Addison-Wesley, Reading, MA. 1993. sis of Past Research with Recommendations for O’Neil, J. M. “Gender-Role Conflict and Strain in Modifications and Future Research,” Journal of Men's Lives: Implications for Psychiatrists, Consumer Research, (15), 1988, pp. 325-343. Psychologists, and Other Human-Service Provi- Sichel, D. E. The Computer Revolution: An Econo- ders,” in Men in Transition: Theory and Therapy, mic Perspective, The Brookings Institution, K. Solomon and N. B. Levy (eds.), Plenum, New Washington, DC, 1997. York, 1982, pp. 5-44. Skitka, L. J., and Maslach, C. “Gender as Oliver, R. L., and Bearden W. O. “Crossover Schematic Category: A Role Construct Effects in the Theory of Reasoned Action: A Approach,” Social Behavior and Personality Moderating Influence Attempt,” Journal of Con- (24:1), 1996, pp. 53-74. sumer Research (12), 1985, pp. 324-340. MIS Quarterly Vol. 24 No. 1/March 2000 135 Venkatesh & Morris/Gender in Technology Acceptance and Usage Sparks, P. “Attitude Towards Food: Applying, Venkatesh, V., and Davis, F. D. “A Model of the Assessing, and Extending the Theory of Planned Perceived Ease of Use Development and Test,” Behavior,” in The Social Psychology of Health Decision Sciences (27:3), 1996, pp. 451-481. and Safety: European Perspectives, D. R. Rutter Venkatesh, V., and Davis, F. D. “A Theoretical and L. Quine (eds.), Aldershot, Avebury, Extension of the Technology Acceptance Model: England, 1994, pp. 25-46. Four Longitudinal Studies,” Management Stein, A. H., and Bailey, M. M. “The Socialization of Science, forthcoming.. Achievement Orientation in Females,” Psycho- Venkatesh, V., Morris, M. G., and Ackerman, P. L. logical Bulletin (80), 1973, pp. 345-366. “A Longitudinal Investigation of Gender Dif- Subramanian, G. H. “A Replication of Perceived ferences in Individual Technology Adoption Deci- Usefulness and Perceived Ease of Use Measure- sion Making Processes,” Organizational Behavior ment,” Decision Sciences (25:5/6), 1994, pp. , forthcoming.. and Human Decision Processes 863-874. Venkatesh, V., and Speier, C. “Comptuer Tech- Szajna, B. “Software Evaluation and Choice: nology Training in the Workplace: A Longitudinal Predictive Validation of the Technology Accep- Investigation of the Effect of Mood,” Organiza- tance Instrument,” MIS Quarterly (18), 1994, pp. tional Behavior and Human Decision Processes 319-324. (79), 1999, pp. 1-28. Szajna, B. “Empirical Evaluation of the Revised Warshaw, P. R. “A New Model for Predicting Technology Acceptance Model,” Management Behavioral Intentions: An Alternative to Fishbein,” Science (42), 1996, pp. 85-92. Journal of Marketing Research (17), 1980, Tashakkori, A. “Gender, Ethnicity, and the Struc- pp.153-172. ture of Self-Esteem: An Attitude Theory Weiner, L. R. Digital Woes: Why We Should Not Approach,” Journal of Social Psychology (133), Depend on Software, Addison-Wesley, Reading, 1993, pp. 479-488. MA. 1993. Tashakkori, A., and Thompson, V. D. “Race Dif- Weissman, M. M., Bland, R., Joyce, P. R., ferences in Self-Perception and Locus of Control Newman, S., Wells, J. E., and Witchen, H. U. During Adolescence and Early Adulthood: “Sex Differences in Rates of Depression: Cross- Methodological Implications,” Genetic, Social, National Perspectives,” Journal of Affective and General Psychology Monographs (117), Disorders (29), 1993, pp. 77-84. 1991, pp. 133-152. Weller, L., Shlomi, A., and Zimont, G. “Birth Order, Taylor, M. C., and Hall, J. A. “Psychological Sex, and Occupational Interest,” Journal of Androgyny: Theories, Methods, and Conclu- Vocational Behavior (8), 1976, pp. 45-50. sions,” Psychological Bulletin (92), 1982, pp. Wilder, G., Macki, D., and Cooper, J. “Gender and 347-366. Computers: Two Surveys of Computer-Related Taylor, S., and Todd, P. A. “Assessing IT Usage: Attitudes,” Sex Roles (13), 1985, pp.215-228 The Role of Prior Experience,” MIS Quarterly Williams, J. E., and Best, D. L. Measuring Sex (19:4), 1995b, 561-570. , Sage, Stereotypes: A Thirty-Nation Study Taylor, S., and Todd, P. A. “Understanding Infor- Beverly Hills, CA, 1982. mation Technology Usage: A Test of Competing Models,” Information Systems Research (6), 1995a, pp.144-176. About the Authors Venkatesh, V. “Creation of Favorable User Perceptions: Exploring the Role of Intrinsic Viswanath Venkatesh is an assistant professor Motivation,” MIS Quarterly (23:2), June 1999, pp. in Decision and Information Technologies in the 239-260. Robert H. Smith School of Business at the Uni- Venkatesh, V. “Determinants of Perceived Ease of versity of Maryland at College Park. He was Use: Integrating Control, Intrinsic Motivation, and named the school’s first Tyser Fellow in 1999. He Emotion Into the Technology Acceptance Model,” received his Ph.D. in Information and Decision Information Systems Research, forthcoming. Sciences from the University of Minnesota in 136 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage 1998. His research focuses on understanding Michael G. Morris is an assistant professor of user acceptance of computer and information Information Systems Management at the Air Force technologies in the workplace and homes. In Institute of Technology, Wright-Patterson AFB, addition, he has a keen interest, from a research Ohio. He received his Ph.D. in Management and teaching perspective, in developing effective Information Systems from Indiana University in methods of user training. His research has been 1996. His research interests center around socio- published (or is forthcoming) in Management cognitive aspects of human response to Science, MIS Quarterly, Information Systems information technology, including technology Research, Organizational Behavior and Human acceptance and usability evaluation. His research Decision Processes, Decision Sciences, Interna- has been published (or is forthcoming) in tional Journal of Human-Computer Studies and Organizational Behavior and Human Decision Personnel Psychology. He was a recipient of the Processes, Decision Sciences, IEEE Software, Smith School’s Teaching Innovation Award in International Journal of Human-Computer Studies 1998. and Personnel Psychology. Appendix A Questionnaire Items Intention to Use Assuming I had access to the system, I intend to use it. Given that I had access to the system, I predict that I would use it. Perceived Usefulness Using the system improves my performance in my job. Using the system in my job increases my productivity. Using the system enhances my effectiveness in my job. I find the system to be useful in my job. Perceived Ease of Use My intention with the system is clear and understandable. Interacting with the system does not require a lot of my mental effort. I find the system to be easy to use. I find it easy to get the system to do what I want it to do. Subjective Norm People who influence my behavior think that I should use the system. People who are important to me think that I should use the system. MIS Quarterly Vol. 24 No. 1/March 2000 137 Venkatesh & Morris/Gender in Technology Acceptance and Usage Gender: Female Male Educational Level: Some high school or less Some college Graduated high school Graduated college Vocational/technical school Post-graduate study Annual Individual Income: Less than $20,000 $60,000 – $69,999 (Before Taxes) $20,000 – $29,999 $70,000 – $79,999 $30,000 – $39,999 $80,000 – $89,999 $40,000 – $49,999 $90,000 – $99,999 $50,000 – $59,999 $100,000 or more Position: Executive/Top Management Administrative/Clerical Middle Management Technical Supervisory Other: _______________________ (please specify) Prior Computer Experience How many years of experience do you have using computers in general? Note: Intention to use, perceived usefulness, perceived ease of use, and subjective norm were measured using a seven-point Likert scale. 138 MIS Quarterly Vol. 24 No. 1/March 2000 Venkatesh & Morris/Gender in Technology Acceptance and Usage Appendix B Measurement Model Estimation Factor Structure Matrix U1 0.92 0.09 0.12 0.07 U2 0.89 0.20 0.19 0.12 U3 0.88 0.15 0.21 0.19 U4 0.95 0.11 0.11 0.04 EOU1 0.13 0.88 0.10 0.21 EOU2 0.02 0.90 0.09 0.22 EOU3 0.14 0.85 0.21 0.03 EOU4 0.09 0.93 0.12 0.07 SN1 0.24 0.07 0.81 0.19 SN2 0.22 0.14 0.83 0.09 BI1 0.27 0.19 0.11 0.87 BI2 0.25 0.16 0.16 0.81 U1 through U4: Perceived Usefulness items EOU1 through EOU4: Perceived Ease of Use items SN1 through SN4: Subjective Norm items BI1 through BI2: Behavioral Intention items Reliability and Discriminant Validity Coefficients ICR 1 2 3 4 Perceived Usefulness 0.93 .91 Perceived Ease of Use 0.92 .22* .88 Subjective Norm 0.85 .37*** .20* .82 Behavioral Intention 0.88 .49*** .30*** .34 .84 Note: Diagonal elements are the square root of the shared variance between the constructs and their measures. Off-diagonal elements are the correlations between the different constructs. ICR = Internal Consistency Reliability * p < .05; *** p < .001 MIS Quarterly Vol. 24 No. 1/March 2000 139
Journal
MIS Quarterly – Unpaywall
Published: Mar 1, 2000