INTRODUCTION

What leads individuals to explore new uses of information technology (IT)? Recent research has examined if either individual differences or social context influences individuals' willingness to adopt and explore IT. Research has linked individual differences to ease of use of IT, an established correlate of technology use (3). In addition, studies have linked the interaction of the work environment and task attributes to technology use (14). However, less research has examined relationships among individual differences, the work environment, and innovative uses of IT.

This paper seeks to extend our understanding of the links between individual differences, situational effects, and innovativeness with IT. Marketing research has found that cultural dimensions influence consumer innovativeness (31). In addition, creativity research has identified situational effects, such as perceived overload, as mediators of personality's influence on innovativeness. We extend these streams of research by examining whether cultural dimensions' effects on PIIT are mediated by perceived overload. Specifically our research questions are:

1. Do cultural dispositions influence willingness to innovate with IT?

2. Do situational effects mediate cultural dispositions' influence on willingness to innovate with IT?

The paper proceeds as follows. First, we develop a theoretical foundation and articulate a research model linking cultural dispositions, situational effects, and innovativeness. Next, using partial least squares (PLS), a structural equation modeling technique, we test the research model. Then we discuss results. The paper concludes with implications for research and practice.

PERSONAL INNOVATIVENESS

Innovativeness may be conceptualized as either a trait or a state. One stream of research suggests that innovativeness reflects innate predispositions to engage in exploratory behaviors (25). From this view, individuals will attempt to innovate regardless of the environment. In addition to innate dispositions, an alternate approach suggests that social context influences innovativeness (15). It suggests that social influences constrain or affect individuals' willingness to try out or engage in exploratory behavior. This paper adopts the latter position and posits that individual differences and the social context influence individuals' willingness to innovate with IT.

Willingness to innovate is a function of personality and perceptions of the environment. At a general level, innovation is thought to be a function of an individual's tolerance for risk (6). If willing to take risks, individuals are more likely to innovate. Across situations, personality's influence varies with individuals' ability to realize target behaviors (13, 23). If they perceive constraints such as a lack of resources or negative outcomes such as punishment, individuals may not express an interest in exploratory or creative behavior. By carefully identifying the target behavior's domain and salient influences, this theory suggests that we may better understand how personality and the environment influence an individuals' willingness to engage in an innovative behavior (2, 13).

In the domain of information technology, personal innovativeness (PIIT) is "the willingness of an individual to try out any new information technology" (4, p. 206). Individuals who are more willing to change will be more likely to express a willingness to innovate with IT. Individuals who are less willing to change will be more likely to require external "reinforcement" before trying an IT innovation (23). In addition, when individuals perceive that they lack resources or skills, they may be less likely to report an integration to use or engage in exploratory behavior with an IT (9, 26). Because external and internal differences may influence expressions of PIIT, this study models it as a disposition that may vary with personality and environmental influences.

NATIONAL CULTURE AND INNOVATIVENESS

Individual dispositions measured by cultural dimensions may influence personal innovativeness with IT. Culture has been defined as "the collective programming of the mind that distinguishes the members of one category of people from those of another" (18, p. 5). Cultural differences have been linked to the diffusion of IT innovations (for example see 32). Hofstede (16, 18) identifies four dimensions of national culture: Uncertainty avoidance, individualism/collectivism, power distance, and masculinity/femininity. These dimensions shape cognitive processes that influence perceptions, motivation, and behavior (10). Recent research suggests that uncertainty avoidance and individualism/collectivism influence consumers' innovativeness (31). Even though these dimensions are frequently used to broadly characterize cultures, research suggests that there may be a substantial variation in cultural dimensions' influence on individuals' perceptions and behavior in a given culture (10). Adler (1) has noted that cultural norms are a group average and individuals' within a culture will exhibit variation around the average. As a result, it is useful to consider the influence of cultural dimensions on individuals' PIIT. The following paragraphs develop hypotheses linking cultural dimensions to individuals' personal innovativeness with information technology.

Uncertainty avoidance may influence individuals' recognition of, and therefore, responses to the opportunities to innovate or explore new uses of IT. Uncertainty avoidance refers to the extent to which individuals avoid ambiguous situations (18). People with low levels of uncertainty avoidance tend to view the unknown as exciting and are more likely to try out new ideas. When high in uncertainty avoidance, people tend to deem ambiguous or novel situations as threatening or risky and consequently are more likely to reject novel ideas (32). As a result, they may identify new applications of IT as more risky and more threatening than sticking to existing patterns of IT usage. Hence:

Hypothesis 1: Uncertainty avoidance will be a negative correlate of personal innovativeness with IT.

Power distance is the extent to which an individual accepts large differentials of power and inequality (18). Individuals high in power distance will be more likely to show greater respect to authority and are less likely to challenge existing knowledge systems. When power distance is large, hierarchical symbols, such as distribution of power, rewards, privileges, and opportunities, are emphasized and unequal distribution of power and rewards are deemed as normal. Since performance and individual input are not the basis for distributing rewards, individuals may be unwilling to innovate because they may not be rewarded for identifying successful new applications of technology. In addition, less powerful members of a high power distance society may feel that they have fewer resources or opportunities necessary to innovate and therefore, over time, be less prone to innovate with an IT. Hence:

Hypothesis 2: Power distance will be a negative correlate of personal innovativeness with IT.

Individualism/collectivism refers to the relationship between individuals and social institutions (31). People who express high levels of collectivist values draw on group membership and participation in traditions as a primary component of their identity. Since innovation frequently disrupts the status quo, highly "collectivist" individuals may be less willing to innovate. At the other end of the continuum, people who express high levels of individualism emphasize the fulfillment of personal values and needs over and above those of groups. Basic social values emphasize personal initiation and achievement. Individuals who exhibit high levels of self-confidence, self-reliance, and bravado are admired and encouraged. Willingness to innovate with IT involves a tendency to initiate new behaviors, independently of others. It is reasonable to expect that such predispositions will be grater in people high on individualism. Indirect support for this reasoning was provided by Steenkamp, Hofstede, and Wedel (31), who found that consumers in individualistic cultures were more likely to buy new products and shift consumption patterns. Hence:

Hypothesis 3: Individualism will be a positive correlate of personal innovativeness with IT.

Masculinity/femininity refers to the degree to which individuals may be characterized by assertiveness or nurturance (18). Highly masculine individuals emphasize the acquisition of money and status, and the achievement of visible and symbolic organizational rewards. They may be characterized as more assertive, confident, and performance oriented. In contrast, highly feminine individuals emphasize quality of life and other less tangible outcomes, such as interpersonal relationships, protection of environment, and caring for others. Research suggests that attributes of masculinity such as assertiveness or self-confidence are positive correlates of innovativeness (22). In addition, because IT innovation is frequently associated with economic success, masculinity may be a precursor to willingness to innovate (21). Because masculine individuals may perceive economic benefits and express more confidence, we posit that "highly masculine" individuals may be more willing to innovate with IT. Hence:

Hypothesis 4: Masculinity will be a positive correlate of personal innovativeness with IT.

THE WORK ENVIRONMENT AND INNOVATIVENESS

While recognizing links between individual differences and domain-specific innovation, the theory suggests that the environment mediates personality's influence on individuals' willingness to innovate. Midgley and Downing (25) argue that simplistic trait-behavior models fail to capture the complexity inherent in innovation processes. They assert that situational effects mediate the influence of traits on expressed innovativeness (pp. 228-230). In an empirical examination of personality, the situation and computer use, Foxall and Bhate (12) found that situational factors exerted a strong effect on overall computer use. Although research cannot ignore personality, they suggest that research focus on identifying "situational effects which hinder or promote ... innovativeness" (12, p. 195).

In this study, we consider the influence of overload, a perception of the work environment, on innovativeness with information technology. Overload refers to individuals' belief that a lack of resources prevents completing an assigned task (28). Qualitative overload exists when individuals believe they lack the capability or skill level required to complete a task. When they report qualitative overload, individuals may believe shortfalls in personal skills or abilities limit their ability to innovate with IT, even though they may have adequate skills. Quantitative overload refers to individuals' belief that limitations imposed by their environment such as time or access to a resource prevent their engaging in a task. When experiencing quantitative overload, individuals may report that they lack the resources required to innovate with IT. In either case, individuals' perceived overload reduces their innovativeness with IT. Hence:

Hypothesis 5a. Qualitative overload will be a negative correlate of personal innovativeness with IT.

Hypothesis 5b. Quantitative overload will be a negative correlate of personal innovativeness with IT.

Mediating Relationship

Creativity research suggests that perceptions of the work environment, such as overload, mediate the relationship between personality and innovativeness (35). According to the componential model of creativity and innovation in organizations, individual differences influence how individuals perceive the environment for creativity (5). The model suggests that individual differences influence perceptions of environmental components. Environmental components fall into two broad categories - impediments and stimulants (5). Impediments refer to obstacles such as overload or a lack of resources. Stimulants refer to the presence of sufficient resources or control over basic job features such as scheduling or methods. This approach suggests that stimulants and obstacles directly influence individuals' willingness to innovate. Prior research has demonstrated that cultural dimensions may be positive or negative correlates of overload (27). This suggests that culture's influence on PUT may be mediated by overload's dimensions.

H6a. Qualitative overload mediates the relationship between cultural dimensions and personal innovativeness with IT.

H6b. Quantitative overload mediates the relationships between cultural dimensions and personal innovativeness with IT.

Figure 1 presents the research model examined in this study.

METHOD

Subjects

Participants were undergraduate business students at a large Southeastern U.S. university. One hundred students participated in the study. Sample characteristics of the data are presented in Table 1.

Procedure

Surveys were administered during regularly scheduled class times. Participants were provided extra credit for responding to the survey.

Measures

Measures were drawn from the management information systems and organizational behavior literatures. Items used to measure each construct are presented in Appendix A. Personal innovativeness with information technology was measured using four items developed by Agarwal and Prasad (2, 4). The four cultural constructs (individualism/collectivism, uncertainty avoidance, power distance, and masculinity/femininity) were measured using scales derived from Hofstede (17). Items were rephrased to reflect cultural changes over the past 20 years. These changes updated language but did not alter the substance of any item (30).1 Overload measures were drawn from the organizational stress literature (7, 29). Overload items were rephrased to direct attention to a specific context, i.e., school, where respondents might be experiencing overload. For all constructs, seven point Likert scales were used, with anchors ranging from strongly agree to strongly disagree.

Preliminary Analysis

Preliminary analysis examined descriptive statistics and response distributions for each measure. An inspection of histograms and scatter-plots indicated that item responses were not normally distributed (33). As a result, the following analysis used techniques robust to non-normal data distributions.

Discriminant and convergent validity are used to suggest that measures of the constructs are distinct and that indicators load on the appropriate construct (24). To evaluate discriminant validity, the AVE may be compared with the square of the correlations among the latent variable (8). The correlation among indicators of a construct should be greater than between a construct and any other construct. The correlation of constructs (see Table 2) demonstrates discriminant validity for the constructs of this study. A second way to evaluate discriminant validity is to examine the factor loadings of each indicator (8). Each indicator should load higher on the construct of interest than on any other variable. Confirmatory factor analysis results, using PLS, confirmed that the observed indicators have adequate discriminant and convergent validity (see Table 3).

RESULTS

Data analysis was conducted using partial least squares (PLS). PLS does not assume multivariate normality among sample distributions (34). In addition, by breaking down models into segments, PLS allows researchers to work with small sample sizes. When determining sample size, theorists suggest that a "rule of thumb" for determining the necessary number of items is ten times the most complex construct's number of indicators or the largest number of paths leading to a latent construct (8). Thus, PLS may be used to test models using small, non-normally distributed datasets. Results are interpreted in two stages - measurement and structural.

Measurement Model

When analyzing the measurement model, researchers evaluate each scale's reliability and validity. To assess internal consistency, PLS researchers typically calculate a block of indicators' internal composite reliability (ICR) and average variance extracted (AVE) (8). Interpreted like a Cronbach's alpha, an ICR of .70 is sufficient for research (11). The AVE measures the variance captured by the indicators relative to measurement error. To use a construct, AVE should be greater than .50 (8). Values reported in Table 3 demonstrate adequate reliability for indicators of the constructs.

Structural Model

Several structural models were estimated. For each model, R^sup 2^ values were calculated for endogenous constructs. Interpreted like multiple regression results, the R^sup 2^ indicates the amount of variance explained by the model (8). Further, using a bootstrapping technique, path estimates and T-statistics were calculated for hypothesized relationships. Analysis proceeded as follows. First, we estimated the direct effects of culture's dimensions on PIIT (H1 to H4). Uncertainty avoidance (H1: p<.01) relationship with PIIT was supported. H2, H3, and H4 were not supported. Next, we added overload's dimensions to the model and re-estimated the model (H5). Results supported hypothesized relationships between qualitative (H5a: p<.01) and quantitative (H5b: p<.05) overload and PIIT. Uncertainty avoidance no longer demonstrated a significant direct effect on PIIT. Then we added paths from culture to overload and tested the mediated relationship. Results provided partial support for H6. Qualitative (H6a: p<.05) and quantitative (H6b: p<-01) overload mediated the relationship from uncertainty avoidance and power distance to personal innovativeness with information technology (Figure 2). Finally, non-significant paths were dropped and a reduced model was estimated. Figure 3 presents the reduced model results.

DISCUSSION

Cultural dimensions were found to have a mediated effect on personal innovativeness with IT. Uncertainty avoidance and power distance had a positive significant influence on both quantitative and qualitative overload. Qualitative and quantitative overload demonstrated a negative significant influence on PIIT. These findings suggest that the work environment mediates cultures influence on PIIT.

Findings have important implications for research and practice. In terms of research, this study extends the nomological net surrounding personal innovativeness with information technology. Although our theoretical view of PIIT differs, this study provides additional evidence of the validity of the innovativeness measures developed by Agarwal and Prasad (2, 3). By linking culture to innovativeness, this study yields insight into enduring influences on individuals' willingness to experiment or explore new uses of information technology. It suggests that individuals who are high in uncertainty avoidance and power distance may be less likely to be willing to innovate or experiment with information technology. By examining culture's direct effect on overload and mediated effect on PIIT, we extend our understanding of the pattern of relationships leading to willingness to innovate with IT. When compared to the work environment, cultural dispositions appear to be relatively distal antecedents to innovativeness with IT.

By linking overload to innovativeness, this study identifies an avenue for managerial action to encourage innovation with IT. Prior research has demonstrated that managerial support influences acceptance and exploration of an IT (23). If managers seek to encourage innovation, our findings suggest that they would be well served to pay attention to the amount and nature of work assigned to employees. Within the domain of organizational behavior, a large body of research has identified means to mitigate the influence of Stressors on employee performance (20). By implementing programs designed to reduce actual or perceived overload, firms may increase employees' willingness to innovate and reap the associated benefits.

Limitations

The primary limitation of this research is the sample. Although it examines the influence of culture, data were collected using students in one country. As a result, this study has limited generalizability to non-students or across countries. However, by controlling for broad cultural differences and the context of data collection, we increased the internal validity of the results. For respondents in this sample, we can say that culture's influence on PIIT was mediated by overload. Results should be regarded as preliminary evidence of culture's influence on innovativeness. Further work should examine the influence of culture and the work environment on innovativeness across different types of individuals and cultures.

Although data were collected in a single country, the U.S., this should not detract from the results found, but strengthen them. If data had been collected in a number of countries greater variation in cultural scores would most likely be found, which could lead to increased levels of significance. Secondly, given the manner in which the hypotheses were stated, respondents from multiple countries were not needed. Examining variation in cultural dimensions within one country was sufficient to provide initial evidence of the model's utility.

Another limitation of this study concerns internal validity. Measures were gathered through self-reports at a single point in times. James, Gent, Hater, and Corey (19) suggest that common method variance is a concern when there appears to be a systematic inflation in the correlation of constructs matrix. Examination of the matrixes in Table 2 demonstrates that correlations varied across constructs. Thus, common method variance does not appear to be a significant flaw in this study.

CONCLUSION

This study examined the patterns of relationships between dimensions of culture, overload, and personal innovativeness with IT. Dimensions of culture and overload were operationalized and tested to determine their relationship with PIIT. Based on data collected from 100 U.S. college students, evidence supported an indirect relationship from two dimensions of culture (i.e., uncertainty avoidance and power distance) and a direct relationship from both quantitative and qualitative overload to PIIT. From a research standpoint, this study extends the nomological net surrounding PIIT. From a practitioner's standpoint, the findings suggest that managers may mitigate the influence of cultural dispositions on PIIT through careful management of the work environment.