Printer Friendly

Why do Students not Major in MIS? An Application of the Theory of Planned Behavior.

1. INTRODUCTION

Recent data (Bureau of Labor Statistics, 2018) shows that the number of IT-related jobs in the U.S. (including computer and information research scientists, software developers, systems analysts, and information security analysts) is projected to grow "13 percent from 2016 to 2026, faster than the average for all occupations." This continued rise in demand for IT-related professionals is attributed to "greater emphasis on cloud computing, the collection and storage of big data, and information security" (p. 1), areas directly relevant to MIS professionals. However recent data also indicates a large and growing number of unfulfilled IT-related jobs. According to a recent USA Today report, 627,000 jobs remain open in U.S. technology-related fields (Swartz, 2017), yet the number of Computer Science and MIS Bachelors and Masters graduates in the previous year was 116,388 (National Center for Education Statistics, 2017). Applications and opportunities for analysis of big data continue to grow and will impact firms around the world. Thus, the lack of workers to fill the information technology worker pipeline is a challenge academicians and practitioners still face.

Although studies that address MIS enrollment issues abound, the manner in which relevant career-deciding factors are collected has been dominated by studies employing surveys based on literature-supplied factors, with fewer studies eliciting factors. A focus of this study is the use of established theory and instruments to elicit and analyze factors that impact MIS enrollment decisions. The use of an inductive approach to identifying important factors complements the body of deductive approaches used in many previous studies.

We use the Theory of Planned Behavior to complete an elicitation study identifying salient behavioral, normative, and control beliefs of college students concerning majoring in MIS. We begin by reviewing extant literature, followed by a discussion of the research model and hypotheses, research methodology, results, and implications. Note that throughout the paper we use the acronym MIS; however, MIS includes IS.

2. LITERATURE REVIEW

2.1 Supplied and Elicited Career Choice Factors

The largest number of studies examining MIS career choice have identified factors from previous literature which are then presented to study participants. In this deductive approach, the researchers ask participants to rate the importance of the given factors in considering an MIS major, and the highest-rated factors overall are deemed the most important (Weinberger, 2004; Walstrom et al., 2008; Hogan and Li, 2011; Li, Zhang, and Zheng, 2014; Snyder and Slauson, 2014).

Only a few studies have elicited factors important to the MIS major decision. Joshi and Kuhn (2011) examined the image students hold of the MIS major by eliciting their descriptions of a "typical MIS professional." The images gleaned from their elicitation were ones commonly found in previous studies, such as "geeky and nerdy" and "lacking social skills." Lee and Lee (2006) interviewed 21 business students to uncover factors important to them and that encourage or inhibit them in choosing their major. The interviews yielded two factors unreported by previous studies: career flexibility and promotional effort, the latter representing "efforts to advertise the major." In one of the few studies to examine factors important to non-U.S. students, Rouibah (2012) elicited factors by surveying 195 students in Kuwait, asking what factors led the student to major in MIS if the student was an MIS major, and why MIS was not attractive to the student if he/she was not majoring in MIS. Rouibah found that intellectual curiosity was a motivator for some Kuwaiti students who decided to major in MIS. Scott et al. (2009) examined factors that encourage or discourage students to major in MIS by eliciting perceptions of the MIS major from 31 students in focus group settings. A key finding of the study was that a lack of understanding about the MIS major contributes to inaccurate perceptions of MIS regarding job scope and variety of work.

Eliciting beliefs can supplement our understanding of why students may or may not major in MIS; however, there are some limitations with some elicitation-based studies. One limitation is that the methods of elicitation tend to vary across studies. Lee and Lee (2006) and Scott et al. (2009) each used focus groups to elicit important factors but with different sample populations; Rouibah (2012) and Joshi and Kuhn (2011) elicited factors using researcher-developed, open-ended questions. The variation in methods is useful for discovering factors, but poses challenges in replicating results and ensuring no method-bias in the resulting factors. Second, protocols used to categorize elicited factors are not always reported or, if reported, vary across the studies. This can call into question the reliability of the factors elicited. Finally, when new factors are discovered, the relative importance of the factors to choosing an MIS career are not often examined.

2.2 Use of Theory in MIS Career Choice Research

When theory has been used to guide the identification and understanding of MIS career choice factors, two theories have been used most often: social cognitive career theory (SCCT) and the theory of reasoned action (TRA).

2.2.1 SCCT. The SCCT is derived in large part from Bandura's (1986) social cognitive theory (SCT) of human development which suggests three mechanisms that form the core of one's development: observed behavior, personal factors (cognitive and affective), and environment (social and physical). These factors are said to interact with each other to facilitate learning and determine the ongoing behavior and "life paths" taken by individuals. Bandura (1986) asserts that self-efficacy beliefs are at the very core of the personal factors component of the SCT and, when combined with other theory factors, exerts influence over which activities are actively pursued.

Lent, Brown, and Hackett (1994) applied the SCT to the study of career choices by individuals, focusing on the SCT's personal factors. The resulting theory was the SCCT. Personal factors emphasized in the SCCT include self-efficacy beliefs, outcome expectations, and goal representations. Several MIS career choice studies have used SCCT to identify and frame factors important to choosing the MIS major (Akbulut and Looney, 2007; Looney and Akbulut, 2007; Koch and Trower, 2011; Akbulut-Bailey, 2012). Looney and Akbulut (2007), Akbulut and Looney (2007), and Akbulut-Bailey (2012) supplied self-efficacy and outcome expectation-related factors to study participants who then rated their perceptions of those factors relative to majoring in MIS. The studies found self-efficacy and outcome expectations to be significant in determining students' interest towards majoring in MIS. However, the studies did not report which specific factors provided the most explanatory value in shaping MIS major outcome expectations or in forming self-efficacy.

2.2.2 TRA and TPB. When it comes to the study of behavioral intentions, the TRA is one of the most commonly-applied theories across a variety of disciplines. The TRA (Fishbein and Ajzen, 1975) attempts to explain how we form intentions to perform specific behaviors. The intentions are influenced by our attitudes toward performing the behavior of interest as well as subjective norms or expectations of other people.

Ajzen (1985) expanded the TRA to account for behaviors that are (or are perceived to be) non-volitional in nature. In describing the need to expand the TRA, Ajzen (1991, pp. 182-183) notes that some behaviors depend
   at least to some degree on such non-motivational
   factors as availability of requisite opportunities and
   resources (e.g., time, money, skills, cooperation of
   others ...) ... these factors represent actual control over
   the behavior ... [but] of greater psychological interest
   than actual control ... is the perception of behavioral
   control ... [which] refers to people's perception of the
   ease or difficulty of performing the behavior of interest.


The addition of the Perceived Behavioral Control construct to the TRA forms the theory of planned behavior (TPB). A further aspect of control beliefs is the type of control being examined. Internal control beliefs represent the degree to which individuals view themselves as skillful enough to perform a behavior of interest (similar in concept to self-efficacy (Bandura, 1977)), versus external control beliefs which capture the extent to which an individual believes factors such as resources or time constraints would help or inhibit their behavior (Ajzen and Madden, 1986).

The Ajzen (1991) protocol for studying human behavior using the TRA/TPB calls for eliciting the most salient beliefs one holds about a behavior under study, as opposed to providing intuitively-determined beliefs. Ajzen notes that elicited beliefs tend to be more closely associated with global measures of attitude than intuitively-determined beliefs (p. 192).

3. RESEARCH MODEL AND HYPOTHESES

Numerous studies have used the TRA or TPB to explain behaviors related to health, safety, classroom behavior, consumer behavior, sports, relationships, college organizations, and social and environmental issues. In the MIS context, TRA and TPB have been used in dozens of studies examining a variety of behaviors including IT adoption (Harrison, Mykytyn, and Riemenschneider, 1997; Grandon and Mykytyn, 2004; Pavlou and Fygenson, 2006) and IT acceptance/use (Davis, Bagozzi, and Warshaw, 1989; Mathieson, 1991; Taylor and Todd, 1995). When it comes to career choice, the TPB in particular has been effective in understanding intentions to major in entrepreneurship (Sieger and Monsen, 2015), physics and media studies (Taylor, 2015), engineering (Mishkin et al., 2016), and careers in general (Amani and Mkumbo, 2016). In a review of 27 studies considering MIS career choices, only 9 of those studies utilized the TRA and TPB to explain the selection of MIS as a major. In one of the earliest studies on MIS career choice based upon the TRA, Trower, Willis, and Dorsett (1994) followed the Ajzen (1991) protocol in eliciting salient factors that influence attitude and subjective norms. Among the key findings of the study was a significant difference in the degree to which intended majors versus intended non-majors felt the MIS major provided an opportunity to graduate with a useful skill and provided a balance between business and technology (Trower, Willis, and Dorsett, 1994). The study found that intended MIS majors weighted these two factors significantly more highly than intended non-majors. Most subsequent TRA-based studies of MIS career choice have been based on work by Zhang (2007). Zhang (2007) reviewed previous literature on choosing college majors and proposed a set of behavioral and normative beliefs that would influence one's attitude and subjective norms towards majoring in MIS. Zhang categorized the supplied beliefs and included them on a survey distributed to undergraduate business students. Subsequent analysis of survey data found support for both attitudes towards majoring in MIS as well as social norms regarding majoring in MIS as being significant influences to the intention to major in MIS.

While the TRA has demonstrated utility in understanding the influencers in the decision to major in MIS, many of the TRA-based studies have been mostly deductive in nature, beginning with beliefs observed in prior research. Further, the number of studies utilizing the TPB in MIS career choice research is smaller than the number of SCCT and TRA-based studies which limits our understanding of whether control-related factors may influence the decision to major in MIS. Our research seeks to determine whether the use of an elicitation-based TPB approach can add to our understanding of the decision to major in MIS.

As previously stated, the TRA (Fishbein and Ajzen, 1975) posits that the attitudes and subjective norms of an individual influence the intentions of that individual to perform a certain behavior, and those intentions influence the actual behavior. In 1985, Ajzen expanded the TRA and labelled this expansion the TPB when he added perceived behavioral control as a third antecedent to behavioral intention. In summarizing the TPB, consistent with the TRA, Ajzen (2006) notes that individual attitudes toward a behavior emanate from underlying salient beliefs about outcomes of the behavior, and that social pressures are influenced by salient beliefs about whether others important to the individual believe he/she should perform the behavior. Thus, we hypothesize:

H1: The more positive the beliefs that majoring in MIS produces certain outcomes, weighted by evaluations of the outcomes, the more positive the attitude towards majoring in MIS.

H2: The more positive the beliefs that specific normative referents feel a student should major in MIS, weighted by the student's motivation to comply with the norms, the more positive the subjective norm towards majoring in MIS.

A possible explanation for the dearth of studies using the TPB in this research stream is the assumption that the decision to major in MIS is voluntary; as Zhang (2007) states "the behavior of interest, (declaring the major) is within the students' control and thus completely volitional, [therefore] it is reasonable to assume that perceived behavioral control does not affect the intentions to choose an IS major" (p. 448). However, the broader definition of perceived behavioral control (PBC) as stated in Ajzen and Driver (1991, p. 188) notes that PBC
   deals with the presence or absence of requisite
   resources and opportunities. The more resources and
   opportunities individuals believe they possess, and the
   fewer obstacles or impediments they anticipate, the
   greater should be their perceived control over the
   behavior.


The more obstacles students perceive they face when considering a major in MIS, the less personal control they may feel over the decision. Therefore, we examine the following hypothesis:

H3: The more positive the beliefs about resources required to major in MIS, weighted by perceptions of the power of these resources, the more positive the perceived behavioral control towards majoring in MIS.

Ajzen (2006) summarizes the core constructs of the TPB and their impacts on behavioral intention. Consistent with the TRA, the TPB suggests that individuals will more likely intend to perform a behavior if they have a positive perception (attitude) toward the behavior and if they perceive favorable views from others regarding performing the behavior (subjective norm). The significance of these core constructs to the decision to major in MIS has been confirmed in prior MIS research by Zhang (2007) and others (Kuechler, McLeod, and Simkin, 2009; Croasdell, McLeod, and Simkin, 2011; Downey, McGaughey, and Roach, 2011; Kumar and Kumar 2013). Hence, we hypothesize the following:

H4: The more positive the attitude about majoring in MIS, the greater the intention to major in MIS.

H5: The more positive the level of subjective norm towards the MIS major, the greater the intention to major in MIS.

In a study examining MIS career choice using a modification of the TPB framework, Ferratt et al. (2010) elicited behavioral beliefs via a survey with open-ended questions devised by the researchers. The study elicited beliefs not previously identified in MIS career choice research such as the ability to have a practical application of MIS coursework. However, the strength and contributions of the elicited factors in predicting intentions to major in MIS were not examined. Heinze and Hu (2009) studied students' choice of IT as a major using a model that integrates the TPB with the SCCT. The beliefs were not elicited or weighted; however, the PBC construct was found to be significantly related to the intention to pursue an IT major, highlighting the potential of the TPB in explaining the MIS career choice. Their finding is consistent with Ajzen (1991) who suggests that the greater one perceives his/her ease of performing a behavior, the stronger his/her intention to perform the behavior should be, leading to the following hypothesis:

H6: The more positive the perception of control over the ability to major in MIS, the greater the intention to major in MIS.

Both the TRA and the TPB indicate that a central factor in predicting actual behavior and the immediate antecedent to actual behavior is the individual's intention (readiness) to perform that behavior, and that the stronger the intention to engage in the behavior, the more likely the actual behavior will be achieved (Ajzen, 1991, 2006). Therefore, we hypothesize:

H7: The more positive the intention to major in MIS, the greater the likelihood of actually majoring in MIS.

To summarize, the decision to major in MIS has been widely-studied. For those utilizing a theory-driven approach, few empirically examined all theory components using protocols described for the theory. Our study addresses this gap by examining the decision to major in MIS using the TPB and its measurement protocols, with the goal of identifying and explaining factors that best influence the decision to major in MIS. The use of the TPB allows us to examine a richer set of factors which include the possibility of perceived internal and external barriers to majoring in MIS. We elicit salient factors from students, as salient beliefs that students hold about majoring in MIS may have changed over time. Figure 1 depicts the proposed research model for this study.

4. METHODOLOGY

Data was collected in two phases. In the first phase, we elicited salient beliefs concerning the study of MIS in college. In the second phase, we conducted the full study using survey questions based on established TPB protocols including elicited beliefs from the first phase.

4.1 Phase I--Elicitation

We elicited beliefs regarding majoring in MIS from a population of 150 students from a university in the southwest region of the United States as well as a university in Zambia, both of which are private universities. We chose this mixture of universities in order to get a broader representation of student beliefs about majoring in MIS. Details of the elicitation study are reported in Chipidza, Green, and Riemenschneider (2016a); a brief summary follows.

Elicitation survey questions were drawn from prior TPB research and consistent with protocols suggested by Ajzen and colleagues (Ajzen and Fishbein, 1980; Ajzen and Driver, 1991), with modifications to suit the context of our study. In total, 136 students completed the survey (response rate of 90%) with 110 usable responses in the final analysis. Of the 110 usable responses, 52% of the sample was female, 43% was classified as juniors or seniors, and 97% was from the U.S. university. Survey responses were coded using content analysis protocols consistent with the approach explicated in Hsieh and Shannon (2005). Two researchers independently coded the responses, with an overall inter-rater agreement rate of 90%. Cohen's kappa ranged from 61% to 100% for the different questions, indicating good to excellent agreement (Fleiss, 1981). Descriptive statistics of elicited beliefs are shown in Table 1. We utilized the "Top 10" rule to select salient beliefs in the behavioral and normative categories which yielded 10 behavioral and 5 normative beliefs (Ajzen and Fishbein, 1980; Sutton et al., 2003). There was an overlap in elicited control and behavioral beliefs. To resolve, we selected those salient control beliefs that were not already covered in the behavioral category for inclusion in the full study, resulting in five control beliefs. Table 2 shows the salient beliefs.

Using the Ajzen (1991) approach to eliciting salient beliefs, we found it interesting that some elicited beliefs overlap between Behavioral Beliefs and Control Beliefs, specifically beliefs expressing "Lack of Interest" and availability of "Job Opportunities." Regarding "Lack of Interest," the overlap between belief constructs may suggest that students view "Lack of Interest" as both an attitude (e.g., "no interest in MIS career") and a perceived obstacle that they feel is difficult or costly to change (e.g., "do not enjoy MIS," "not technology-inclined," "no interest in problem-solving"). Similarly, "Job Opportunities" surfaced as both positive attitudes about majoring in MIS ("competitive advantage"" and "high salary") as well as beliefs about the MIS major that would make it easy or less costly to major in MIS (e.g., "job placement"). These results suggest the possibility of measurement and/or theoretical overlaps between behavioral and control beliefs for this particular set of elicited beliefs. According to Fishbein and Ajzen (2011), this overlap of beliefs is highly probable in empirical research.

4.2 Phase II--Full Study

4.2.1 Survey Development. Phase II of our study collected data to study the decision to major in MIS. Items to measure beliefs about majoring in MIS, as well as items that provide for direct assessment of attitude (ATT), subjective norm (SN), perceived behavioral control (PBC), and intention (INT) constructs closely followed guidelines published in Ajzen (2002), modified to fit the context of majoring in MIS. Items to measure the behavioral, normative, and control beliefs were created from the most common items elicited in Phase I. For example, the first behavioral belief provided in Table 2 is "competitive advantage" and the resulting question on the Phase II survey from this belief was "Majoring in MIS in college would give me a competitive advantage when seeking out a job." The first normative belief from Table 2 is "family" and the resulting question on the Phase II survey is "My family thinks that I should major in MIS." All of the measurement items follow a similar pattern and are included in the Appendix.

The salient beliefs garnered from Phase I formed the basis of the indirect measurement questions to assess participants' behavioral, normative, and control beliefs towards majoring in MIS. Consistent with Ajzen (2002), two questions are asked for each behavioral belief: the degree of the respondent's belief about majoring in MIS and the importance attached to that belief. Similarly, each normative belief generates two questions: the degree to which the participant feels others expect him/her to major in MIS and the participant's motivation to comply with the normative referent's expectations. Last, each control belief is also assessed through two questions. The composite of the strength and power of each belief provides a measure of the belief's impact in facilitating or impeding a participant's intention to major in MIS.

4.2.2 Data Collection. We conducted a pilot study in spring 2015 with 150 student participants from the private university in the southwest U.S. Analysis of the 147 usable responses revealed that "motivation to comply" questions about normative beliefs had been omitted. After correcting the survey to add these questions, we began the full study by emailing a link to the electronic survey to professors at the same two universities that participated in the elicitation study as well as a public university in the southeast U.S. Professors disseminated the link to their students offering a class credit incentive to complete the survey. Most students had taken or were taking a required introductory IT course. Three hundred thirty five students responded to the survey; cases with at least 50% missing values were removed leaving 319 usable cases. Table 3 shows sample characteristics.

To assess construct validity, we performed a factor analysis of the indicators. Since the primary theoretical constructs are expected to correlate highly, we included their items in a principal axis factor analysis with an oblique rotation. The items loaded highest on their theoretical constructs except for item SN3 ("The people in my life whose opinion I value would approve of me majoring in MIS in college"), which did not load above 0.3 on any factor and was therefore dropped. The factor analysis of the indicators is not included due to space constraints. Belief factors in TPB are also expected to highly correlate (Ajzen, 2002). Therefore, we conducted a principal axis factor analysis with an oblique rotation in order to determine the underlying factor structure of behavioral, normative, and control beliefs shown in Table 4.

All normative belief items loaded on one factor as expected. Two control beliefs loaded on a second factor. The two items (CB2 and CB4) measured students' perceptions of their interest and self-efficacy in technology. However, behavioral beliefs loaded in an unexpected way. A third factor was composed of five items: two related to behavioral belief items and three related to control belief items. An examination of these five items revealed they were related through their focus on MIS job-related characteristics; for example, BB1 assessed student perceptions of job market competitiveness whilst CB3 assessed students' perceptions of the ease of finding a job in MIS. The overlap between behavioral and control beliefs is not new nor unprecedented and is empirically highly probable (Fishbein and Ajzen, 2011). Kraft et al. (2005) note that some measures of behavioral control beliefs may indeed reflect the experiential aspect of attitude, which may have been a factor with the aforementioned three control belief items. A fourth factor was composed of two items related to students' intrinsic motivations to major in MIS. The items, BB5 and BB10, measured participants' perceptions of how satisfying and interesting majoring in MIS would be. Items BB4 and BB8 loaded highest on a fifth factor capturing beliefs about technical skills. The former item assessed participants' beliefs that majoring in MIS would help them acquire technical skills while the latter measured the belief that MIS graduates would have to maintain pace with new technological developments. Items BB2, BB6, and BB9 loaded together on a sixth factor and tapped aspects of cost incurred by majoring in MIS. Specifically, the items measure the difficulty, time commitment, and social cost perceived when majoring in MIS.

The preceding four factors form a primary layer of constructs for a second-order behavioral beliefs (BB) construct. Figure 2 depicts the hierarchical component model for the newly-developed BB construct. Only item BB3 ("If I major in MIS in college, I will have to write a lot of programming code") did not load above 0.3 on any factor and was deleted.

5. RESULTS

To analyze relationships among the constructs in the research model, we chose the component-based Partial Least Squares (PLS) method for three reasons. First, the final model was reasonably complex, employing a mixture of formative and reflective constructs, and a second-order construct. Second, we were unable to guarantee normality of the collected data a priori. Third, the belief constructs obtained from the factor analysis are not well tested; hence this research is somewhat exploratory in nature (Urbach and Ahlemann, 2010; Hair et al., 2013).

The sections below report the results of the PLS-based analyses. The analyses include all data collected from the three universities described earlier. While the small number of responses from the U.S. public university and the Zambian university did not allow us to do comparative analyses to determine if models based on data from each separate university were consistent, when analyzing the research model using the U.S.-based private data alone (89% of the sample), we found the results to be consistent with the results reported below.

5.1 Measurement Model

For each participant, indicator scores for the belief constructs were calculated as a product of belief strength and the importance attached to that belief as described earlier. These indicator formulations highlight an important strength of TPB research: a salient belief only motivates an individual to perform a stated behavior if that individual attaches importance to that belief.

Each item of the belief components captures a different aspect of the underlying TPB construct. The normative items capture student perceptions of friends, family, advisors, teachers, and IT professionals' opinions about majoring in MIS. These measures will not necessarily correlate since a student's parent may hold a different opinion concerning MIS than an academic advisor would. Indeed, belief constructs are not expected to have high internal consistency (Harrison, Mykytyn, and Riemenschneider, 1997; Ajzen, 2002;). Salient beliefs concerning MIS collectively explain students' behavioral, normative, and control beliefs in the MIS context. According to MacKenzie, Podsakoff, and Jarvis (2005), if a construct's indicators are its defining characteristics, then a formative indicator measurement model should be specified. Past TRA research has formatively modeled belief constructs (Zhang, 2007; Downey, McGaughey, and Roach, 2011; Hennessy, Bleakley, and Fishbein, 2012; Kumar and Kumar, 2013). As such, belief constructs can be appropriately modeled formatively, indicating that changes in the indicators lead to changes in the construct (Petter, Straub, and Rai, 2007). On the other hand, items constituting the major TPB constructs (ATT, SN, PBC, and INT) are interchangeable in that removing a single item should not alter the conceptual meaning of the construct; hence, these constructs are modeled reflectively.

We assessed formative constructs on the basis of their collinearity (Hair et al., 2013). The variance inflation factor (VIF) statistic for all formative indicators ranged from 1.01 to 2.24, below the maximum threshold of 3.3 beyond which collinearity becomes a problem when estimating the path model (Diamantopoulos and Siguaw, 2006). In calculating outer weights of the indicators, NNB3, NNB4, and NCB3 did not have significant weights, indicating they did not contribute significantly to their respective formative constructs in relative terms. Consequently, we assessed the absolute contributions of the formative indicators to the constructs by looking at their loadings. NNB3 (loading = -0.08, p = 0.74) and NNB4 (loading = 0.07, p = 0.25) did not significantly load on their construct. The non-significant contributions of NNB3 and NNB4 showed that they did not contribute towards the formation of the normative beliefs construct; thus, we deleted them.

Reflective constructs are analyzed on the criteria of internal consistency, convergent validity, and discriminant validity (Roldan and Sanchez-Franco, 2012). Tables 5 and 6 summarize results of these analyses. The model's reflective constructs exceed the Hair et al. (2013) benchmarks for demonstrating high internal consistency. Indicator reliability and average variance extracted (AVE) are used to assess the convergent validity of each reflective construct. Except for item PBC1, reflective item loadings ranged from 0.70 to 0.97, meeting the requirement for indicator reliability. We considered PBC1 for removal from the model; however, its deletion did not raise the composite reliability or AVE above the threshold values. Consequently, we retained PBC1 in the model (Hair et al., 2013). AVE values for reflective constructs ranged from 0.56 to 0.92 demonstrating adequate convergent validity. In determining discriminant validity, the squared correlation between two constructs must be less than either of their individual AVEs (Fornell and Larker, 1981). As Table 6 shows, this condition is met for all the model's reflective constructs. Cross loadings can also be used to evaluate the model's discriminant validity. Analysis of the cross loadings for the indicators in the model showed acceptable levels of discriminant validity.

5.2 Tests for the Second-Order Behavioral Beliefs Construct

Our factor analysis revealed that behavioral beliefs fall into four categories: technical skills, job factors, intrinsic motivations, and cost. Thus, behavioral beliefs is a second-order construct composed of four first-order constructs (Figure 2). Each first-order construct captures a different aspect of the behavioral belief construct. Therefore, this hierarchical component model is modeled as a formative-formative type, indicating that the relationship between the first- and second-order constructs is formative, and each construct is measured by formative indicators (MacKenzie, Podsakoff, and Jarvis, 2005). We used a repeated indicators approach in that the second-order construct used all the indicators of the first-order constructs in its measurement (Lohmoller, 1988).

There were significant relationships between three first-order constructs (job factors, cost, and intrinsic motivations) and the second-order construct (see Figure 3). However, the relationship between the technical skills construct and the second-order construct was non-significant. According to Mackenzie, Podsakoff, and Podsakoff (2011), lower-order constructs with non-significant relationships are candidates for elimination only if they exhibit multicollinearity problems. The VIF for the Technical Skills to Behavioral Beliefs path was 1.27, well below the benchmark of 3.3 (Diamantopoulos and Siguaw, 2006). This suggests there were no multicollinearity problems with the technical skills construct; thus, the construct was retained in the model. Finally, we tested whether the second-order construct mediates the relationships between each first-order construct and ATT; our test indicates that BB indeed fully mediates the relationships - the influences of technical skills considerations, job factors, cost factors, and intrinsic motivations on attitude towards studying MIS are fully explained by the second-order BB construct.

5.3 Structural Model

The structural model is assessed on the basis of strength of the relationships between the constructs. The initial model included controls for age, gender, ethnicity, college level, and socioeconomic status; these are omitted from subsequent models as none of the control variables had a significant relationship with behavioral intention. Figure 3 depicts our structural model. Similar to the measures for assessing formative indicators, VIF values for each predictor construct should not exceed the benchmark of 5 (Hair et al., 2013). Our model's VIF values range from 1.00 to 1.63 indicating no significant collinearity among the predictor constructs. Four intervening variables are in the research model: attitude (ATT), subjective norm (SN), perceived behavioral control (PBC), and intention to major in MIS (INT). Hayes (2009) argues that bootstrapping, as opposed to the Sobel test or causal steps approach, must be used to investigate mediating variables. We selected a sample of 5,000 with a confidence interval of 95% as bootstrapping parameters. T values exceeding 1.96 for a given path indicate a significant relationship between the two constructs linked by the path. Running the PLS algorithm reveals that all paths were significant except the paths between technical skills and behavioral beliefs and between PBC and intention. Table 7 summarizes the results of hypothesized relationships.

The relationship between behavioral beliefs and attitude towards majoring in MIS is empirically validated ([beta] = 0.74, p = 0.000), lending support towards H1. Specifically, positive behavioral beliefs are associated with favorable attitudes towards majoring in MIS. Similarly, normative beliefs positively correlate with subjective norms concerning majoring in MIS, suggesting that strong normative beliefs are associated with high perceived social pressure ([beta] = 0.39, p = 0.000); H2 is supported by our results. H3 is also supported; control beliefs positively correlate with perceived behavioral control ([beta] = 0.45, p = 0.000).

Attitude and subjective norms towards MIS are positively related to behavioral intention ([beta] = 0.41, p = 0.000) and ([beta] = 0.38, p = 0.000) respectively, lending support towards both H4 and H5. However, the relationship between PBC and intention is non-significant ([beta] = 0.03, p = 0.50); hence, H6 is not supported. Finally, H7 is supported; behavioral intention is positively associated with actual behavior--participants that expressed intention to major in MIS were more likely to do so than participants that did not intend to major in MIS ([beta] = 0.74, p = 0.000).

To assess the predictive value of the model, the coefficient of determination, [R.sup.2], is employed. The model explains 51% of the variance in the intention to major in MIS and 55% of the variance in actual behavior. The size of the effect of ATT on INT is 0.22 and that of SN on INT is 0.21 (Table 7), representing medium effects for the predictor variables on INT (Cohen, 1988).

In addition, we evaluated the predictive relevance of the model using the [Q.sup.2] statistic (Stone, 1974). [Q.sup.2] values greater than 0 indicate predictive relevance for the construct (Hair et al., 2013). [Q.sup.2] values ranged from 0.06 (SN) to 0.47 (actual enrollment behavior), indicating that the model has predictive relevance for its endogenous constructs. Specifically, the model correctly predicts indicator values in reflective and single-item constructs.

6. DISCUSSION AND POST-HOC ANALYSIS

This study sought to understand the strongest influencers of students' decisions to major (or avoid majoring) in MIS and the beliefs that shape these influencers. The TPB-based approach found strongest support for the influences of students' attitude towards majoring in MIS, and the expectations of other people important to the student, in this order. Results reported in Figure 3 give us additional insights into the beliefs that most strongly shape these influencers.

Intrinsic motivators, such as interest and the potential for a rewarding and satisfying experience, proved to be the strongest influencer of a students' attitude about majoring in MIS, followed closely by job-related aspects of MIS, such as competitive advantage, salary, support structure, and ease of finding a job that MIS careers offer. Analysis also revealed that the perception of the MIS major as posing high costs to students can lessen their positive attitude towards majoring in MIS. Examples of this include difficulty of classes in the major, number of additional classes required for the major, and the perceptions of limited social interaction in college and beyond with a career in MIS.

Perceived social pressures were also found to significantly predict a student's intention to major in MIS. Of particular interest is the influence of IT professionals, which has the most significant impact on a students' intention to major in MIS, followed by influences of friends and family. Interestingly, student participants did not perceive social pressures to major in MIS from high school teachers nor from high school or college academic advisors as demonstrated by the lack of significant weights and loadings of these indicators.

The perception of the ease or difficulty of majoring in MIS lacked significance in predicting intention to major in MIS. This finding contradicts findings of Heinze and Hu (2009), as well as the original proposition by Ajzen (1991), that PBC influences behavioral intention. Therefore, we conducted further analyses on the PBC construct to determine if its effect could be observed in other ways. The TPB proposes that PBC may directly influence actual behavior in addition to the proposed indirect effect through intention. We tested that relationship in a post-hoc analysis and found the relationship was statistically significant ([beta] = 0.08, p = 0.03) at p = 0.05. However, the size of the effect ([f.sup.2] = 0.01) fell below the required threshold for it to be characterized as even a small effect (Cohen, 1988).

Further consideration of PBC is that the present study incorporated items measuring both internal and external control aspects of PBC. Ajzen (2002) acknowledged growing questions over the structure of the PBC component of TPB, noting findings of differing impacts of these two aspects of PBC. Citing a meta-analysis of the PBC construct by Cheung and Chan (2000), Ajzen reported studies finding internal (self-efficacy) PBC items significantly explaining behavioral intentions, while external (controllability) PBC items were more significant in explaining actual behavior (Ajzen, 2002). Our operationalization combining both components of PBC suggested higher levels of PBC are associated with a higher likelihood of actual enrollment in MIS giving further credence to the possibility of PBC being multi-dimensional with differing impacts on the decision to enroll in MIS. Thus, it is clear that more work remains in conceptualizing the PBC construct in the context of majoring in MIS.

Finally, we conducted two-tailed independent t-test analyses to examine variations in attitudes towards MIS and perceived control over the decision to major in MIS according to whether the students are MIS majors or not. Results of these analyses are unsurprising; compared to non-majors, MIS majors had significantly more positive attitudes, more positive perceptions of subjective norms, and greater perceived control over the decision to major in MIS. Further, MIS majors more positively perceived the job market for MIS professionals than non-majors. These findings underscore the characteristics that differentiate students that elect to major in MIS from those that do not.

In sum, our analyses confirm the utility of TPB constructs in predicting behavioral intention to major in MIS. The results suggest implications to practice for influencing these important factors and implications to research about the structure and nature of influence of the TPB constructs when it comes to predicting decisions to major in MIS.

7. IMPLICATIONS

Our study has implications for both MIS theory and practice. We will address both the theoretical and the practitioner views in an interwoven fashion as they interrelate to each other.

First, based on a full implementation of the TPB methodology (Ajzen, 1991), we have demonstrated that the TPB provides a useful framework for understanding and predicting why students major in MIS. In particular, our study confirms that attitudes towards majoring in MIS and normative beliefs about whether to major in MIS are significant predictors of the intention to major in MIS. The application of this theory to studying MIS enrollment is relevant from a global perspective as the need to grow the international workforce and fill the MIS worker pipeline continues to be an area of economic concern. Given the strong demand and projections for MIS and ICT-related workers (Manyika, 2017; Bureau of Labor Statistics, 2018), our research can inform MIS practitioners of currently-held beliefs about majoring in MIS, which can inform their efforts to increase the pipeline of MIS workers.

Second, our study extends the behavioral beliefs dimension of the TPB by uncovering four sub-constructs of behavioral beliefs that emerge when determining intentions to major in MIS. These sub-dimensions allow researchers to analyze and understand the MIS career decision phenomena with more granularity and depth than prior TPB-based career choice studies. Of these sub-constructs, intrinsic motivations had the strongest influence on behavioral beliefs followed by job factors, cost, and technical skills. It is insightful to know that these students were most influenced by the interest, satisfaction, and reward of the major followed by the ability to secure a job after graduation and earn a good salary. From a practitioner standpoint, to further enhance the interest, satisfaction, and reward of the major, video testimonials emphasizing successful MIS majors could be developed and uploaded on university websites. MIS departments and professional recruiters could work together to create these videos highlighting job satisfaction and rewards of recent hires. Both MIS professionals and faculty should emphasize job availability and good salaries in promotional marketing pieces catered to students in junior high and high school, as well as college freshmen.

Third, in this study, the PBC construct was not significant in predicting intention to major in MIS, thus raising the question of applicability of PBC for this behavior of interest. However we further found PBC significant in predicting actual behavior, albeit with a small effect size. Thus, another contribution of this research is to raise the need for further research regarding the PBC construct and its direct or indirect influence on majoring in MIS.

Fourth, our use of TPB-based elicitation uncovered behavioral, normative, and control beliefs about majoring in MIS not typically uncovered in previous studies. These factors include the opportunity to stay abreast of rapid-changing technology; the availability of a support system of teachers, family, and friends when majoring in MIS; and the strong influence of IT professionals on decisions to major in MIS. While these factors had varying levels of contributions to our research model, the emergence of these factors give support for the periodic use of TPB-based elicitation methods to uncover new MIS career decision factors that may emerge over time (Taylor, 2015). In addition, these factors are prescriptive to faculty and practitioners alike. Faculty can offer a support system to the MIS majors through professional groups, such as student AITP chapters and faculty/student mentoring programs. Additionally, faculty can expose MIS majors to conferences such as the Grace Hopper Conference by offering scholarships and sponsorship for student participation.

As the impact of IT professionals in influencing students to major in MIS was a significant contributor to our research model, industry practitioners are encouraged to increase mentorship programs that encourage and support students in their early years of education. MIS department faculty can work with professional advisory board members to arrange mentorship programs and offer events for interaction between the students and MIS professionals. Firm sponsored events such as pre-game tailgating offer opportunities for student and professional interaction in an informal and fun environment and can help mitigate perceptions of limited social interaction in the MIS field. In addition, recent research supports the necessity of IT professionals' influence specifically in encouraging more women and minorities to pursue IT careers (Wang, 2017).

The lack of influence of high school teachers and advisors found in our study supports the findings of other studies that have noted the lack of adequate coverage of MIS careers at the secondary education level (Walstrom et al., 2008; Downey, McGaughey, and Roach, 2009; Chipidza, Green, and Riemenschneider, 2016b). Clayton, Beekhuyzen, and Nielsen (2012) advocate interventions designed to influence middle-school girls to major in ICT fields, noting that there are few existing influences in the middle and high school environments supporting a culture of majoring in ICT fields. Our findings suggest that interventions at the secondary and higher education levels can be effective by emphasizing intrinsic motivators and MIS job characteristics as ways to positively influence attitudes towards majoring in MIS. Interventions should also address the key cost-related concerns that may cause students to form negative attitudes towards majoring in MIS, including the feeling that MIS classes are too difficult, that majoring in MIS would limit their social interaction, and that majoring in MIS would require too many additional classes. Efforts such as ensuring consistent supply of tutors and offering access to tutorials can ensure potential MIS majors that ample resources are available for success.

8. LIMITATIONS AND FUTURE RESEARCH

As with any empirical study, there are limitations to the study. First, student participants provided self-reported judgments regarding behavioral intention; however, unlike other studies, actual behaviors were observed and reported. Since student subjects were the population of interest, the critique often associated with using students as subjects is not warranted for this study. Additionally, we conducted a Harman's one factor test as a counter to the limitation of common method bias, and we found no bias.

Second, we have utilized the elicitation approach suggested by Ajzen when researching behavioral intentions using the TRA/TPB. However, this approach does constrain us by TPB-based assumptions. Thus, some beliefs not related to TPB constructs could be excluded, such as fears, trust, past experiences, etc.

Future research should consider comparisons across different countries, including residents in low and middle-income countries such as some African countries and India, as well as more economically-developed and developing countries such as Ireland, the Netherlands, and Poland. This extension to other countries with a variety of cultural backgrounds could allow for additional insight into cultural differences of student selection or not of MIS as a major. Additional data from a variety of countries would also allow for comparison across ethnicity and gender in future research.

Finally, an area not as frequently explored in MIS research is the attraction to MIS due to its applicability to other industries that may be of more "interest" to students such as sports, health care, accounting, sociology etc. Future research in this area, including the exploration of gender-based attractions to these related disciplines, is suggested.

9. CONCLUSION

The current study has demonstrated that the TPB provides a useful framework for understanding and predicting why students major in MIS. The availability of a support system for students majoring in MIS, the opportunity to stay well-informed of rapid-changing technology, and the influence of MIS professionals are factors not identified in prior research that should be considered as we work to increase the pool of MIS majors globally. Second, we decomposed behavioral beliefs into four sub-constructs: job characteristics, intrinsic motivations, technical skills, and cost factors, allowing us to analyze the data with more granularity and depth than in extant studies. Third, based on the insignificant findings of PBC influencing intention, we raise the question of applicability of the PBC construct for predicting intention to major in MIS. Additional research is needed to validate or refute this finding.

10. REFERENCES

Ajzen, I. (1985). From Intentions to Actions: A Theory of Planned Behavior. In J. Kuhl and J. Beckmann (Eds.), Action-Control: From Cognition to Behavior (pp. 11-39). Heidelberg: Springer.

Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and Human Decision Processes, 50(2), 179-211.

Ajzen, I. (2002). Constructing a TPB Questionnaire: Conceptual and Methodological Considerations. Retrieved from http://chuang.epage.au.edu.tw/ezfiles/168/1168/attach/20/pt a_41176_7688352_57138.pdf.

Ajzen, I. (2006). Constructing a Theory of Planned Behavior Questionnaire. Unpublished Manuscript.

Ajzen, I. & Driver, B. L. (1991). Prediction of Leisure Participation from Behavioral, Normative, and Control Beliefs: An Application of the Theory of Planned Behavior. Leisure Sciences, 13(3), 185-204.

Ajzen, I. & Fishbein, M. (1980). Understanding Attitudes and Predicting Social Behaviour. Englewood Cliffs, NJ: Prentice-Hall.

Ajzen, I. & Madden, T. J. (1986). Prediction of Goal-Directed Behavior: Attitudes, Intentions, and Perceived Behavioral Control. Journal of Experimental Social Psychology, 22(5), 453-474.

Akbulut, A. & Looney, C. (2007). Inspiring Students to Pursue Computing Degrees. Communications of the ACM, 50(10), 67-71.

Akbulut-Bailey, A. (2012). Improving IS Enrollment Choices: The Role of Social Support. Journal of Information Systems Education, 23(3), 259-270.

Amani, J. & Mkumbo, K. (2016). Predictors of Career Intentions among Undergraduate Students in Tanzania. Journal of Education and Human Development, 5(3), 106-115.

Bandura, A. (1977). Self-Efficacy: Toward a Unifying Theory of Behavioral Change. Psychological Review, 84(2), 191-215.

Bandura, A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall.

Bureau of Labor Statistics. (2018). U.S. Department of Labor, Occupational Outlook Handbook, Computer and Information Technology Occupations. Retrieved from https://www.bls.gov/ooh/computer-and-informationtechnology/home.htm.

Cheung, S. F. & Chan, D. K. S. (2000). The Role of Perceived Behavioral Control in Predicting Human Behavior: A Meta-Analytic Review of Studies on the Theory of Planned Behavior. Unpublished Manuscript, Chinese University of Hong Kong.

Chipidza, W., Green, G., & Riemenschneider, C. (2016a). Salient Beliefs in Majoring in Management Information Systems: An Elicitation Study. Information Systems Education Journal, 14(4), 69-80.

Chipidza, W., Green, G., & Riemenschneider, C. (2016b). Choosing Technology Majors: What You Don't Know can Influence You. Journal of Computer Information Systems, 59(1), 1-14.

Clayton, K., Beekhuyzen, J., & Nielsen, S. (2012). Now I Know what ICT can do for Me! Information Systems Journal, 22(5), 375-390.

Cohen, J. (1988). Statistical Power Analysis for the Behavioral Science (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum Associates.

Croasdell, D., McLeod, A., & Simkin, M. (2011). Why Don't More Women Major in Information Systems? Information Technology and People, 24(2), 158-183.

Davis, F., Bagozzi, R., & Warshaw, P. (1989). User Acceptance of Computer Technology: A Comparison of Two Theoretical Models. Management Science, 35(8), 982-1003.

Diamantopoulos, A. & Siguaw, J. A. (2006). Formative Versus Reflective Indicators in Organizational Measure Development: A Comparison and Empirical Illustration. British Journal of Management, 17(4), 263-282.

Downey, J., McGaughey, R., & Roach, D. (2009). MIS Versus Computer Science: An Empirical Comparison of the Influences on the Students' Choice of Major. Journal of Information Systems Education, 20(3), 357-368.

Downey, J., McGaughey, R., & Roach, D. (2011). Attitudes and Influences toward Choosing a Business Major: The Case of Information Systems. Journal of Information Technology Education, 10(1), 231-251.

Ferratt, T. W., Hall, S. R., Prasad, J., & Wynn, Jr., D. (2010). Choosing Management Information Systems as a Major: Understanding the smiFactors for MIS. Communications of the Association for Information Systems, 27(1), 265-284.

Fishbein, M. & Ajzen, I. (1975). Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley.

Fishbein, M. & Ajzen, I. (2011). Predicting and Changing Behavior: The Reasoned Action Approach. New York, NY: Taylor & Francis.

Fleiss, J. L. (1981). The Measurement of Interrater Agreement. In Statistical Methods for Rates and Proportions (2nd ed.), pp. 212-236. New York: John Wiley.

Fornell, C. & Larker, D. F. (1981). Evaluating Structural Equation Models with Unobservable Variables and Measurement Error. Journal of Marketing Research, 18(1), 39-50.

Grandon, E. & Mykytyn, Jr., P. (2004). Theory-Based Instrumentation to Measure the Intention to Use Electronic Commerce in Small and Medium Sized Businesses. Journal of Computer Information Systems, 44(3), 44-57.

Hair, J. F., Hult, G. T. M., Ringle, C., & Sarstedt, M. (2013). A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM). Thousand Oaks, CA: Sage Publications.

Harrison, D., Mykytyn, P., & Riemenschneider, C. (1997). Executive Decisions about Adoption of Information Technology in Small Business: Theory and Empirical Tests. Information Systems Research, 8(2), 171-195.

Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical Mediation Analysis in the New Millennium. Communication Monographs, 76(4), 408-420.

Heinze, N. & Hu, Q. (2009). Why College Undergraduates Choose IT: A Multi-Theoretical Perspective. European Journal of Information Systems, 18(5), 462-475.

Hennessy, M., Bleakley, A., & Fishbein, M. (2012). Measurement Models for Reasoned Action Theory. The ANNALS of the American Academy of Political and Social Science, 640(1), 42-57.

Hogan, P. & Li, L. (2011). The Perceptions of Business Students Regarding Management Information Systems (MIS) Programs. Journal of Technology Research, 2, 1-8.

Hsieh, H.-F. & Shannon, S. E. (2005). Three Approaches to Qualitative Content Analysis. Qualitative Health Research, 15(9), 1277-1288.

Joshi, K. D. & Kuhn, K. (2011). What Determines Interest in an IS Career? An Application of the Theory of Reasoned Action. Communications of the Association of Information Systems, 29(1), 133-158.

Koch, H. & Trower, J. (2011). How I Became IS: Understanding the Major Decision. Proceedings of the America's Conference on Information Systems, Detroit, MI.

Kraft, P., Rise, J., Sutton, S., & Roysamb, E. (2005). Perceived Difficulty in the Theory of Planned Behavior: Perceived Behavioral Control or Affective Attitude? British Journal of Social Psychology, 44(3), 479-496.

Kuechler, W., McLeod, A., & Simkin, M. (2009). Why Don't More Students Major in IS? Decision Sciences Journal of Innovative Education, 7(2), 463-488.

Kumar, A. & Kumar, P. (2013). An Examination of Factors Influencing Students' Selection of Business Majors Using TRA Framework. Decision Sciences Journal of Innovative Education, 11(1), 77-105.

Lent, R., Brown, S., & Hackett, G. (1994). Toward a Unifying Social Cognitive Theory of Career and Academic Interest, Choice, and Performance. Journal of Vocational Behavior, 45(1), 79-122.

Lee, Y. & Lee, S. J. (2006). The Competitiveness of the Information Systems Major: An Analytic Hierarchy Process. Journal of Information Systems Education, 17(2), 211-221.

Li, L., Zhang, C., & Zheng, G. (2014). Promoting Information Systems Major to Undergraduate Students--A Comprehensive Investigation. Journal of Information Systems Education, 25(3), 211-219.

Lohmoller, J.-B. (1988). The PL S Program System: Latent Variables Path Analysis with Partial Least Squares Estimation. Multivariate Behavioral Research, 23(1), 125-127.

Looney, C. & Akbulut, A. (2007). Combating the IS Enrollment Crisis: The Role of Effective Teachers in Introductory IS Courses. Communications of the Association of Information Systems, 19(1), 781-805.

MacKenzie, S. B., Podsakoff, P. M., & Jarvis, C. B. (2005). The Problem of Measurement Model Misspecification in Behavioral and Organizational Research and Some Recommended Solutions. Journal of Applied Psychology, 90(4), 710-730.

MacKenzie, S. B. Podsakoff, P. M., & Podsakoff, N. P. (2011). Construct Measurement and Validation Procedures in MIS and Behavioral Research: Integrating New and Existing Techniques. MIS Quarterly, 35(2), 293-334.

Manyika, J. (2017). Technology, Jobs, and the Future of Work. McKinsey Global Institute Briefing Note for the Fortune Vatican Forum, December 2016. Retrieved from https://www.mckinsey.com/~/media/McKinsey/Global%20 Themes/Employment%20and%20Growth/Technology%20j obs%20and%20the%20future%20of%20work/MGI-Futureof-Work-Briefing-note-May-2017.ashx.

Mathieson, K. (1991). Predicting User Intentions: Comparing the Technology Acceptance Model with the Theory of Planned Behavior. Information Systems Research, 2(3), 173-191.

Mishkin, H., Wangrowicz, N., Dori, D., & Dori, Y. (2016). Career Choice of Undergraduate Engineering Students. Procedia-Social and Behavioral Sciences, 228, 222-228.

National Center for Education Statistics (NCES). (2017). Digest of Education Statistics. Retrieved from https://nces.ed.gov/programs/digest/d17/tables/dt17_318.30 .asp.

Pavlou, P. & Fygenson, M. (2006). Understanding and Predicting Electronic Commerce Adoption: An Extension of the Theory of Planned Behavior. MIS Quarterly, 30(1), 115-143.

Petter, S., Straub, D., & Rai, A. (2007). Specifying Formative Constructs in Information Systems Research. MIS Quarterly, 31(4), 623-656.

Roldan, J. L., & Sanchez-Franco, M. J. (2012). Variance-Based Structural Equation Modeling: Guidelines for using Partial Least Squares. In M. Mora, O. Gelman, A. Steenkamp, & M. Raisinghani (Eds.), Research Methodologies, Innovations and Philosophies in Software Systems Engineering and Information Systems, (pp. 193-221). Hershey: IGI Global.

Rouibah, K. (2012). Understanding Student Drivers and Obstacles toward MIS Major from the Perspective of an Arab Country: The Case of Kuwait. Issues in Information Systems, 13(2), 58-71.

Scott, C., Fuller, M., MacIndoe, K., & Joshi, K. D. (2009). More than a Bumper Sticker: The Factors Influencing Information Systems Career Choices. Communications of the Association for Information Systems, 24(1), 7-26.

Sieger, P. & Monsen, E. (2015). Founder, Academic, or Employee? A Nuanced Study of Career Choice Intentions. Journal of Small Business Management, 53(1), 30-57.

Snyder, J. & Slauson, G. (2014). Majoring in Information Systems: Reasons Why Students Select (or not) Information Systems as a Major. Information Systems Education Journal, 14(1), 4-12.

Stone, M. (1974). Cross-Validatory Choice and Assessment of Statistical Predictions. Journal of the Royal Statistical Society. SeriesB (Methodological), 36(2), 111-147.

Sutton, S., French, D. P., Hennings, S. J., Mitchell, J., Wareham, N. J., Griffin, S., Hardeman, W., & Kinmonth, A. L. (2003). Eliciting Salient Beliefs in Research on the Theory of Planned Behaviour: The Effect of Question Wording. Current Psychology, 22(3), 234-251.

Swartz, J. (2017). Tech Jobs are Thriving Nationwide - Up to 7.3M. USA Today. Retrieved from https://www.usatoday.com/story/tech/talkingtech/2017/04/0 3/tech-jobs-thriving-nationwide--up-73m/99789502/.

Taylor, R. (2015). Using the Theory of Planned Behavior to Understand Students' Subject Choices in Post-Compulsory Education. Research Papers in Education, 30(2), 214-231.

Taylor, S. & Todd, P. A. (1995). Understanding Information Technology Usage: A Test of Competing Models. Information Systems Research, 6(2), 144-176.

Trower, J., Willis, G. W. K., & Dorsett, D. (1994). An Evaluation of Factors Influencing Intentions to Major in Information Systems. Journal of Information Systems Education, 6(4), 206-211.

Urbach, N., & Ahlemann, F. (2010). Structural equation modeling in information systems research using partial least squares. JITTA: Journal of Information Technology Theory and Application, 11(2), 5-40.

Walstrom, K., Schambach, T., Jones, K., & Crampton, W. (2008). Why are Students not Majoring in Information Systems? Journal of Information Systems Education, 19(1), 43-54.

Wang, J. (2017). Is the U.S. Education System Ready for CS for All? Communications of the ACM, 60(8), 26-28.

Weinberger, C. (2004). Just Ask! Why Surveyed Women Did not Pursue IT Courses or Careers. IEEE Technology and Society Magazine, 23(2), 28-35.

Zhang, W. (2007). Why IS: Understanding Undergraduate Students' Intentions to Choose an Information Systems Major. Journal of Information Systems Education, 18(4), 447-458.
Appendix 1: Descriptive Statistics for TPB Measurement Items

Model        Item   Definition
Variable

Attitude     ATT1   For me, majoring in MIS in college would be
                    good/bad

             ATT2   For me, majoring in MIS in college would be
                    enjoyable/unenjoyable

             ATT3   For me, majoring in MIS in college would be
                    pleasant/unpleasant

             ATT4   For me, majoring in MIS in college would be
                    wise/unwise

             ATT5   For me, majoring in MIS in college would be
                    valuable/worthless

             ATT6   For me, majoring in MIS in college would be
                    interesting/boring

Behavioral   BB1    Majoring in MIS in college would give me a
Beliefs             competitive advantage when seeking out a job

             BB2    If I major in MIS in college, I will find the
                    classes difficult

             BB3    If I major in MIS in college, I will have to write
                    a lot of programming code

             BB4    If I major in MIS in college I will acquire
                    additional technical skills

             BB5    If I major in MIS in college, I will find it
                    interesting

             BB6    If I major in MIS in college, I will have to take
                    many additional classes

             BB7    If I major in MIS in college, I will earn a high
                    salary

             BB8    If I major in MIS in college, I will have to keep
                    pace with rapidly changing technology

             BB9    If I major in MIS in college, I will have limited
                    interaction with people during college or during
                    my professional life

             BB10   If I major in MIS in college, I will find it a
                    rewarding and satisfying experience

Control      CB1    I expect that majoring in MIS will require more
Beliefs             effort compared to other majors

             CB2    I am interested in technology

             CB3    It will be easy to find a job if I major in MIS

             CB4    I am good at technology

             CB5    If I majored in MIS, my teachers, friends and
                    family would support me

Behavioral   INT1   I intend to major in MIS
Intention

             INT2   I will major in MIS

             INT3   I plan to major in MIS

Normative    NB1    My family thinks that I should major in MIS
Beliefs

             NB2    My friends think that I should major in MIS

             NB3    My academic advisors think that I should major in
                    MIS

             NB4    My high school teachers that I should major in MIS

             NB5    IT professionals think that I should major in MIS

Perceived    PBC1   Majoring in MIS is completely up to me

Behavioral   PBC2   For me to major in MIS would be easy

Control      PBC3   For me to major in MIS would be possible

             PBC4   If I wanted to I could be an MIS major

             PBC5   I have the intellectual ability to be an MIS major

Subjective   SN1    Most people who are important to me think I should
Norm                major in MIS in college

             SN2    Most people who are important to me will
                    themselves major in MIS

             SN3    The people in my life whose opinion I value would
                    approve of me majoring in MIS in college

             SN4    The people in my life whose opinion I value will
                    major in MIS

Model        Item    Mean    Standard
Variable                     Deviation

Attitude     ATT1    1.24    1.58

             ATT2    0.31    1.93

             ATT3    0.36    1.81

             ATT4    1.22    1.63

             ATT5    1.71    1.33

             ATT6    0.71    1.92

Behavioral   BB1     1.59    1.30
Beliefs
             BB2     0.80    1.48

             BB3     0.29    1.57

             BB4     1.92    1.27

             BB5     0.45    1.84

             BB6     0.71    1.56

             BB7     1.43    1.29

             BB8     2.09    1.10

             BB9     0.53    1.64

             BB10    0.47    1.65

Control      CB1     0.77    1.40
Beliefs
             CB2     0.62    1.92

             CB3     1.14    1.52

             CB4     0.73    1.62

             CB5     1.80    1.44

Behavioral   INT1    0.88    2.20
Intention
             INT2   -1.01    2.24

             INT3   -1.03    2.27

Normative    NB1     0.07    1.73
Beliefs
             NB2    -0.14   -0.14

             NB3     0.08    1.41

             NB4    -0.08    1.32

             NB5     1.12    1.63

Perceived    PBC1    2.30    1.29
Behavioral
Control      PBC2   -0.33    1.68

             PBC3    1.23    1.72

             PBC4    1.97    1.54

             PBC5    1.82    1.55

Subjective   SN1    -0.11    1.66
Norm
             SN2    -0.96    1.60

             SN3     1.01    1.78

             SN4    -0.92    1.53


AUTHOR BIOGRAPHIES

Wallace Chipidza is an assistant professor in the Center for Information Systems and Technology at Claremont Graduate University where he teaches courses in Software Development and Research Methods. He holds a Ph.D. in Information Systems from Baylor University and an M.S. in Computer Science from the University of Arizona. He mainly researches the dynamics of social networks, software development teams, and ICT for development. His work has been published in journals that include Communications of the Association for Information Systems (CAIS) and the Journal of Computer Information Systems (JCIS). His work has also been included in the proceedings of the International Conference on Information Systems, the Hawaii International Conference on Systems Sciences, and the Americas Conference on Information Systems, among others.

Gina Green is an associate professor of information systems at Baylor University. She earned her Ph.D. in MIS at the University of South Florida, her M.S.E. in Computer Science from the University of Pennsylvania, and her B.S. in Computer Science from Southern University. She has worked professionally as a systems engineer, DBA, and software development manager for IBM. Dr. Green's primary teaching focus is in database and data warehousing courses. Her current research interests include the study of both information and constituency relationship management in nonprofit organizations, cross-cultural social innovation collaborations, and IS education issues. Her research has been published in the Journal of Computer Information Systems, Communications of the ACM, IEEE Software, Information and Software Technology, and others.

Cynthia K. Riemenschneider is associate dean for research and faculty development and professor of information systems in the Hankamer School of Business at Baylor University. She also holds the Helen Ligon Professorship in Information Systems. Her publications have appeared in MIS Quarterly, Information Systems Research, Journal of Management Information Systems, Information Systems Journal, European Journal of Information Systems, IEEE Transactions on Software Engineering, and others. She currently conducts research on IT work force issues, women and minorities in IT, IT professional turnover and turnaway intention, as well as ethical issues surrounding IT use.

Wallace Chipidza

Center for Information Systems and Technology

Claremont Graduate University

Claremont, CA 91711, USA

Wallace.Chipidza@cgu.edu

Gina Green

Cindy Riemenschneider

Information Systems Department

Hankamer School of Business

Baylor University

Waco, TX 76798, USA

Gina_Green@baylor.edu, C_Riemenschneider@baylor.edu

Caption: Figure 1. Research Model

Caption: Figure 2. Hierarchical Component Model for Behavioral Beliefs

Caption: Figure 3. PLS Results for Student Enrollment in MIS Major
Table 1. Descriptive Statistics for Elicited Beliefs (adapted from
Chipidza, Green, and Riemenschneider, 2016a)

Beliefs                           Question         Total   Mean Beliefs
                                                            per Person
                                                               (SD)

Behavioral                   Like or enjoy          263    2.39 (1.64)
(attitude towards majoring   Advantages             384    3.49 (1.41)
in MIS)                      Dislike or hate        206    1.87 (1.72)
                             Disadvantages          252    2.29 (1.78)
Control                      Easy                   180    1.64 (1.39)
(barriers, facilitators)     Difficult              216    1.96 (1.46)
Normative Referents          Individuals, Groups    146    1.42 (1.67)

Beliefs                          Percent of       Cohen's
                                Respondents        Kappa
                             Supplying Three or
                                more Beliefs

Behavioral                          39%            0.86
(attitude towards majoring          77%            0.70
in MIS)                             29%            0.61
                                    39%            0.75
Control                             20%            0.81
(barriers, facilitators)            13%            0.85
Normative Referents                 21%            1.00

Table 2. Behavioral, Normative, and Control Beliefs Included in
Full Study

Belief Type   Belief Description

Behavioral    Job Opportunities (competitive advantage)
              Acquisition of Technical Skills
              Negative Image
              Difficult Classes
              Time-Consuming
              Constantly-Changing Technology
              Lack of Interest
              Job Opportunities (high salary)
              Personally Rewarding
              Difficulty with Programming

Normative     Family
              Friends
              MIS/IT Professionals
              Advisors
              Educators

Control       Too much Additional Effort
              Lack of Interest
              Lack of Ability/Self-Efficacy
              Job Opportunities (job placement)
              Individual Support System

Belief Type    Number of    Percent of
              Respondents   Respondents

Behavioral        92            84%
                  90            82%
                  83            75%
                  82            75%
                  51            46%
                  48            44%
                  47            43%
                  43            39%
                  37            34%
                  31            28%

Normative         24            71%
                  14            41%
                   7            21%
                   5            15%
                   2            6%

Control           47            46%
                  39            38%
                  39            38%
                  20            21%
                  12            12%

Table 3. Demographic Information, Full Study (N=319)

             Category             Frequency   Percent
                                                of
                                              Sample

Gender       Female                  160        50%
             Male                    158        50%
             Unknown                  1         0%

College      Freshman                66         21%
Level        Sophomore               53         17%
             Junior                  106        33%
             Senior                  83         26%
             Fifth Year              10         3%
             Unknown                  1         0%

University   U.S. Private            283        89%
             U.S. Public             20         6%
             Zambian                 13         4%
             Unknown                  3         1%

Age in       18--19                  91         29%
Years        20--21                  147        46%
             22--23                  53         17%
             >23                     27         8%
             Unknown                  1         0%

Race/        African                 52         16%
Ethnicity    American/Black
             Hispanic/Latino         28         9%
             Caucasian or White      209        66%
             American Indian or       1         0%
             Alaska Native
             Asian                   22         7%
             Unknown                  7         2%

Major        Other                   39         12%
             Unspecified             29         9%
             CS, Other IT-
             related major            6         2%
             MIS                     54         17%
             Business Non-MIS
             or Unsure               191        60%

Table 4. Rotated Factor Pattern Matrix: Behavioral Beliefs, Control
Beliefs, and Normative Beliefs

Construct               Indicator   Description

Behavioral Beliefs:     BB1         Majoring in MIS in college would
Job Characteristics                 give me a competitive advantage
                                    when seeking out a job

                        BB7         If I major in MIS in college, I
                                    will earn a high salary

                        CB1         I expect that majoring in MIS will
                                    require more effort compared to
                                    other majors

                        CB3         It will be easy to find a job if I
                                    major in MIS

                        CB5         If I majored in MIS, my teachers,
                                    friends and family would support me

Behavioral Beliefs:     BB5         If I major in MIS in college, I
Intrinsic Motivations               will find it interesting

                        BB10        If I major in MIS in college, I
                                    will find it a rewarding and
                                    satisfying experience

Behavioral Beliefs:     BB4         If I major in MIS in college I will
Technical Skills                    acquire additional technical skills

                        BB8         If I major in MIS in college, I
                                    will have to keep pace with rapidly
                                    changing technology

Behavioral Beliefs:     BB2         If I major in MIS in college, I
Cost Factors                        will find the classes difficult

                        BB6         If I major in MIS in college, I
                                    will have to take many additional
                                    classes

                        BB9         If I major in MIS in college, I
                                    will have limited interaction with
                                    people during college or during my
                                    professional life

Control Beliefs         CB2         I am interested in technology

                        CB4         I am good at technology

Normative Beliefs       NB1         My family thinks that I should
                                    major in MIS

                        NB2         My friends think that I should
                                    major in MIS

                        NB3         My academic advisors think that I
                                    should major in MIS

                        NB4         My high school teachers that I
                                    should major in MIS

                        NB5         IT professionals think that I
                                    should major in MIS

                                      Factor
Construct
                        1      2      3      4      5      6

Behavioral Beliefs:     0.55
Job Characteristics

                        0.64

                        0.52

                        0.61

                        0.50

Behavioral Beliefs:            0.82
Intrinsic Motivations

                               0.73

Behavioral Beliefs:                   0.66
Technical Skills

                                      0.47

Behavioral Beliefs:                          0.42
Cost Factors

                                             0.44

                                             0.51

Control Beliefs                                     0.49

                                                    0.58

Normative Beliefs                                          0.61

                                                           0.56

                                                           0.90

                                                           0.51

                                                           0.38

Table 5. Composite Reliability, Cronbach's Alpha,
and Average Variance Extracted for Reflective Model
Constructs

Construct          Cronbach's Alpha    Composite    Average Variance
                                      Reliability      Extracted

Attitude                 0.92            0.94             0.72
Intention                0.95            0.97             0.92
Perceived
Behavioral               0.77            0.83             0.56
Control
Subjective Norms         0.72            0.82             0.60

Table 6. Correlation Matrix for Model Constructs. (Numbers in bold
are Fornell and Larker Criterion Figures (square root of AVE);
F = Formative Constructs; PLS does not calculate Fornell

Construct                          1       2       3       4

1. Attitude                        0.85#
2. Behavioral Beliefs              0.74    F
3. Control Beliefs                 0.47    0.52    F
4. Cost Beliefs                    0.33    0.45    0.26    F
5. Enrolled in MIS                 -0.49   -0.42   -0.22   -0.17
6. Intention                       0.64    0.61    0.37    0.25
7. Intrinsic Motivation Beliefs    0.71    0.96    0.50    0.29
8. Job Factors Beliefs             0.47    0.64    0.33    0.26
9. Normative Beliefs               0.38    0.43    0.30    0.15
10. PBC                            0.42    0.39    0.45    0.23
11. Subjective Norms               0.56    0.52    0.34    0.26
12. Technical Skills Beliefs       0.28    0.39    0.24    0.16

Construct                          5             6        7      8

1. Attitude
2. Behavioral Beliefs
3. Control Beliefs
4. Cost Beliefs
5. Enrolled in MIS                 Single Item
6. Intention                       -0.74         0.96#
7. Intrinsic Motivation Beliefs    -0.40         0.60     F
8. Job Factors Beliefs             -0.28         0.31     0.45   F
9. Normative Beliefs               -0.27         0.40     0.42   0.31
10. PBC                            -0.30         0.31     0.32   0.38
11. Subjective Norms               -0.41         0.62     0.49   0.35
12. Technical Skills Beliefs       -0.18         0.11     0.37   0.41

Construct                          9      10      11      12

1. Attitude
2. Behavioral Beliefs
3. Control Beliefs
4. Cost Beliefs
5. Enrolled in MIS
6. Intention
7. Intrinsic Motivation Beliefs
8. Job Factors Beliefs
9. Normative Beliefs               F
10. PBC                            0.12   0.75#
11. Subjective Norms               0.39   0.29    0.77#
12. Technical Skills Beliefs       0.19   0.28    0.14    F

Note: Fornell and Larker Criterion Figures (square root of AVE) are
indicated with #

Table 7. Research Hypotheses Results

     Hypothesis

H1   The more positive the beliefs that majoring in MIS produces
     certain outcomes, weighted by evaluations of the outcomes, the
     more positive the attitude towards majoring in MIS.

H2   The more positive the beliefs that specific normative referents
     feel a student should major in MIS, weighted by the student's
     motivation to comply with the norms, the more positive the
     subjective norm towards majoring in MIS.

H3   The more positive the beliefs about resources required to major
     in MIS, weighted by perceptions of the power of these resources,
     the more positive the perceived behavioral control towards
     majoring in MIS.

H4   The more positive the attitude about majoring in MIS, the greater
     the intention to major in MIS.

H5   The more positive the level of subjective norm towards the MIS
     major, the greater the intention to major in MIS.

H6   The more positive the perception of control over the ability to
     major in MIS, the greater the intention to major in MIS.

H7   The more positive the intention to major in MIS, the greater the
     likelihood of majoring in MIS.

                    Path
      Result     Coefficient   Effect Size
                  (p-value)    ([f.sup.2])

H1   Supported   0.74 (0.00)      1.20

H2   Supported   0.39 (0.00)      0.20

H3   Supported   0.45 (0.00)      0.26

H4   Supported   0.41 (0.00)      0.22

H5   Supported   0.38 (0.00)      0.21

H6      Not      0.03 (0.50)      0.00
     supported

H7   Supported   0.74 (0.00)      1.20
COPYRIGHT 2019 Journal of Information Systems Education
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Chipidza, Wallace; Green, Gina; Riemenschneider, Cindy
Publication:Journal of Information Systems Education
Date:Mar 22, 2019
Words:12033
Previous Article:Creating a Request for Proposal for Software for a NonProfit Organization.
Next Article:Developing a Framework to Understand Student Engagement, Team Dynamics, and Learning Outcomes Using ERPsim.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters