Microcomputer Playfulness: Stable or Dynamic Trait?

Business Computer Information Systems
College of Business Administration
University of North Texas
P O Box 305249
Denton, Texas 76203
Phone: (940) 565-3110
Facsimile: (940) 565-4935

Accepted for publication in an upcomping special issue of
The DATA BASE for Advances in Information Systems on "Play and Computers"

Microcomputer Playfulness: Stable or Dynamic Trait?

Previous playfulness research has investigated playfulness as both state and trait phenomena. For example, Webster et al. (1993) examined flow, the state of playfulness in a specific human-computer interaction, while Martocchio & Webster (1992) used a trait-based approach, considering playfulness a characteristic of individuals. This research extends the investigation of playfulness as an individual trait by using a longitudinal study to examine its temporal and situational stability. The Computer Playfulness Scale (Webster & Martocchio, 1992) was administered four times over the course of a five week summer session to students enrolled in a computer-literacy course, once at the beginning of the class and then following completion of three milestones in the course work. The playfulness instrument was assessed for internal consistency, unidimensionality, and temporal and situational stability. The evidence indicates that the measurement is reliable. The primary question of trait stability (stable versus dynamic) was examined in several ways, supporting the conclusion that playfulness is a stable trait. The implications of these findings and suggested further research are discussed.

Keywords: Playfulness, longitudinal study, traits, cognitive playfulness, cognitive spontaneity, Computer Playfulness Scale.

ACM Categories: H.1.2 User/Machine Systems, human factors; J.4 Social and Behavioral Sciences, psychology; K.6.1 Project and People Management, systems development and training

Increasingly, MIS designers are able to add "playful" items to systems. Flying toaster screen savers, Porky Pig’s voicing of audible cues, and desktops constructed in themes tied to Disney characters inhabit a growing number of computers. Moreover, new multimedia capabilities and the advent of virtual reality offer new methods to further increase microcomputer playfulness. Concurrent with these new playfulness-enhancing technologies, system designers have growing abilities to customize and individualize systems. Increasingly sophisticated individual agents have begun to lurk in cyberspace. Application packages and operating systems have almost universally adopted user-adjustable graphical user interfaces (GUIs) which are frequently customizable.

These new capabilities underscore the need to better understand the role of playfulness in system design and training. Information Systems professionals face a critical issue in understanding when playfulness augments the learning or operating of a system, when playfulness may serve as a distraction, and how the appropriate use of playfulness may depend on individual and system differences.

The importance of individual differences in the design and operation of Information Systems can be traced to the earliest frameworks of information systems. For example, "an information system consists of, at least, a PERSON of a certain PSYCHOLOGICAL TYPE . . . " (Mason & Mitroff, 1972, p. 475) is one of the earliest frameworks for defining Information Systems. In addition to other effects psychological types or traits have, individual differences may affect users’ learning about new software; and some researchers perceive a critical need to match training methods to these individual differences (e.g., Bostrom, Olfman, & Sein, 1990).

Over the last ten years, psychologists seeking to explain individual differences in personality and behavior increasingly subscribe to trait theories. Furthermore, the most popular of these psychological trait theories is the five factor model (FFM). This personality model (based on dimensions of Neuroticism, Extraversion, Openness, Agreeableness, and Conscientiousness) is characterized as: "a basic discovery" (McCrae & John, 1992), the basis for the field of personality and individual differences (Buss, 1989), and sufficient to characterize both normal and abnormal behavior (Widiger, 1993). However, despite the general acceptance of trait theory as key in understanding human behavior, there is no generally accepted definition of the term "personality trait." Personality traits are generally thought of as long term predispositions to certain behaviors or attitudes. Two generally accepted conditions of personality traits are temporal stability and cross-situational consistency (e.g., Veenhoven, 1994).

In the MIS literature, traits are defined as static aspects of human information processing characteristics affecting a broad range of variables (Bostrom, Olfman, & Sein, 1990). General traits refer to comparatively stable characteristics of individuals that are relatively invariant to situational stimuli (Webster & Martocchio, 1992). Cognitive traits are based on processing preferences and include cognitive styles (Bostrom, Olfman, & Sein, 1990). The effect of individual traits on computer usage has a rich history in IS literature, including recent work concentrating on computer self-efficacy (e.g., Compeau & Higgins, 1995), computer anxiety (e.g., Fajou, 1996), and conscientiousness (e.g., Stewart, Carson, & Cardy, 1996).

MIS professionals seeking to match both the systems and the training methods for these systems to individual differences should not only consider differences among individuals but also whether these differences are dynamic. In particular, professionals should consider whether users’ attitudes or behaviors might change as they gain exposure to a system. If the individual traits are not stable (either temporally or situationally), the problem of matching these traits to system characteristics becomes decidedly more difficult.

Playfulness is considered a multi-faceted construct, encompassing five dimensions: cognitive spontaneity, social spontaneity, physical spontaneity, manifest joy, and sense of humor (Barnett, 1990; Barnett, 1991; Lieberman, 1977). These five dimensions are illustrated as follows: cognitive spontaneity is the imaginative play of young children and the combinatorial play of creative adults; social spontaneity is the ability to be comfortable in a group setting and to move freely in and out of such a social structure; physical spontaneity is evident in unstructuredplay activities such as jumping rope; manifest joy bears different labels such as pleasure and happiness; and sense of humor results from surprising, incongruous, or novel events, whether the individual is the producer or the consumer (Lieberman, 1977). In recent publications and for this study, cognitive spontaneity in human-computer interactions is considered a surrogate for "cognitive playfulness" (Martocchio & Webster, 1992). Cognitive playfulness has been studied as a trait that influences ease of microcomputer use and resultant learning. "Employees higher in cognitive playfulness demonstrated higher test performance and more positive affective outcomes than those lower in cognitive playfulness" (Martocchio & Webster, 1992, p. 553). In addition, those higher in playfulness are expected to exercise and develop skills through exploratory behaviors (Miller, 1973), resulting in improved performance or increased learning (Martocchio & Webster, 1992).

There are, however, potential drawbacks of playfulness, such as requiring a longer time to complete tasks (Sandelands, 1988), over-involvement (Csikszentmihalyi, 1975), and increased opportunities for non-productive play (Nash, 1990). Organizations must be aware that playfulness may result in wasted time, but it may also result in more effective, more productive, and higher-quality results (Starbuck & Webster, 1991).

The Computer Playfulness Scale (CPS) developed by Webster and Martocchio (1992) is a self-reported instrument. It is designed to measure microcomputer playfulness, a situation-specific individual characteristic which represents the degree of cognitive spontaneity in microcomputer interactions (Webster & Martocchio, 1992). Furthermore, microcomputerplayfulness demonstrates higher predictive efficacy for training effectiveness (learning or understanding), compared to previously utilized computer anxiety and computer attitudes (Webster & Martocchio, 1992). Test-retest reliability has proven strong (correlation .85, p<.001) in previous studies using the CPS (Webster & Martocchio, 1992).

Because of the growing ability to manipulate the playfulness of computer systems and training, the Computer Playfulness Scale measure represents a potentially powerful tool allowing system designers to address the interaction of system and individual playfulness. However, before system designers can accommodate the construct of playfulness, its trait nature must be more fully explored. In particular, this research seeks to establish the temporal stability and situational consistency of the playfulness construct.

The subjects were volunteer undergraduate students enrolled in a computer-literacy course at a moderately large southwestern university and received course credit for their participation. Each of seventy-seven subjects was asked to complete four iterations of Webster and Martocchio’s (1992) Computer Playfulness Survey instrument (see Appendix) over a five week summer session, once at the beginning of the course and again following the completion of three milestones in the course work. The administrations are referred to as: 1) Initial administration, baseline first day of class; 2) Windows, following introduction to and projectcompletion using the Microsoft Windows operating environment; 3) Word, following introduction to and project completion using Microsoft Word; and 4) Excel, following introduction to and project completion using Microsoft Excel.

The playfulness score was determined by adding together (i.e., a linear sum) the responses of each individual for the seven items identified by Webster and Martocchio (1992) as comprising the playfulness construct: spontaneous, unimaginative, flexible, creative, playful, unoriginal, and uninventive. This was done after adjusting for the three items that were reverse-scored, compensating for yea-saying or nay-saying individuals who have a more or less global tendency to agree or disagree (Alreck & Settle, 1995).

Two primary goals of this research were to test the temporal stability and situational consistency of the playfulness construct. Psychologists evaluating the temporal stability of other personality traits have selected periods as short as several days or as long as several years in evaluating trait stability. Since the focus of this research was playfulness during microcomputer training, a five week training period was used. In addition to being similar in length to other trait studies (e.g., Stewart, Carson, & Cardy, 1996), this period meets or exceeds the length of time of training in most industry training environments.

End-user microcomputer training is subject to constraints that make micro-computer playfulness less subject to environmental variation than many other personality constructs. One of the most important variants in computer training is task, more specifically the type of software to be learned. Three of the most common software groups are operating systems, word processing, and spreadsheets (e.g., Jones & Berry, 1995). This research tests across these software groups as cross-situational variables.

Internal consistency for the seven-item playfulness instrument was assessed with Cronbach’s (1951) coefficient alpha, "probably the best estimate of internal consistency" (Crano & Brewer, 1973, p. 230). The results are shown in Table 1 below. Based on the greater than .80 rule-of-thumb (Crano & Brewer, 1973; Nunnally, 1978; Blau, 1988), these coefficients indicate that the seven-item playfulness instrument appears to have high internal consistency. Test-retest reliability was also examined and is discussed under hypothesis testing.

Administration	Cronbach’s Alpha
Initial (n = 60)	.9029
Windows (n = 62)	.8825
Word (n = 60)	.8656
Excel (n = 49)	.9383

Another method for assessing internal consistency is to determine whether items "share only one common focus" (Crano & Brewer, 1973, p. 231). The unidimensionality of the scale was evaluated by means of the factorial validity (Kappelman, 1995) of the seven-item scale using the SPSS/PC+ FACTOR procedure (SPSS, Inc., 1993). Each of the four administrations of the playfulness instrument resulted in all seven items loading on a single factor. The first eigenvalues, percent of variance explained by the first eigenvalue, ratio of the first eigenvalue to the second, and range of factor loadings are shown in Table 2 for each administration. Eigenvalues (3.888 to 5.143) and percent of variance (55.5% to 73.5%) are relatively large for allof the four administrations, indicating a consistently high percentage of variance explained by the first factor. The ratio of the first to the second eigenvalues is also substantial, ranging from 4.260:1 to 7.649:1. The factor loading should attain a minimum of 0.50 (Straub, 1989) to be considered as part of a factor. Each of the administrations surpasses that level on all seven items. Unidimensionality is supported by these results, especially by the large factor loadings.

Administration	Eigenvalue	Percent of Variance	Ratio of First:Second	Range of Factor Loadings
Initial	4.454	63.6	5.643:1	.65400 - .90428
Windows	4.182	59.7	4.377:1	.61597 - .85471
Word	3.888	55.5	4.260:1	.70640 - .81484
Excel	5.143	73.5	7.649:1	.68962 - .88752

Previous research has stated that playfulness is a trait. This study tested the hypothesis that playfulness is a trait and there will be no change in playfulness scores over time or across situations. First this hypothesis was tested by examining the correlations between scores obtained by the same person on multiple administrations of the same instrument (Anastasi, 1988). This is the same statistical procedure used to perform test-retest reliability of instruments. Reliability coefficient (r) values of at least .70 indicate that the results are stable over time (Litwin, 1995). However, caution must be exercised when interpreting these results. Practice effect may falsely inflate the correlations (Litwin, 1995). As individuals become familiar withthe items on a survey, they may simply answer based on their memory of how they answered previously (Litwin, 1995). The length of the instrument, which included at least three other instruments in each administration, was designed in part to minimize this problem.

To further confirm that the learning effect was not a serious threat to the experiment, the authors analyzed the change in variance by individuals across administrations. In the event of a significant learning bias, one would expect decreasing variance as answers became "more pat" (that is, individual’s responses would increasingly mirror the previous set of responses). The data showed a slight increase in variance from the first inter-item variance measure (based on individual differences between administrations 1 and 2) to the last (based on differences between administrations 3 and 4). Although this analysis does not preclude a learning effect between the first and second administrations, in the opinion of the authors if such an effect was significant it would likely increase in subsequent administrations. Thus, learning effect did not appear to be a significant threat to this investigation.

These reliability coefficients between administrations of the same instrument represent correlations between the linear sums. Since it may be possible for two consecutive administrations to exhibit little difference while cumulative differences over several administrations may indicate a substantial difference, each result was compared with all other administrations (see Table 3 below). The playfulness scores remained substantively invariant across the four administrations, supporting the stable trait characterization of the playfulness construct.

Administration	Correlation	Significance
Initial with Windows (n = 52)	.842	.000
Initial with Word (n = 50)	.767	.000
Initial with Excel (n = 41)	.669	.000
Windows with Word (n = 54)	.822	.000
Windows with Excel (n = 45)	.783	.000
Word with Excel (n = 44)	.901	.000

The correlations appeared to weaken between non-consecutive administrations over time, bringing into question either the test-retest reliability over short time periods or raising the possibility that playfulness is dynamic and not a stable trait. Comparisons were made of the means and standard deviations (see Table 4 below) using a one-way analysis of variance (ANOVA), and no significant difference in playfulness was found for any of the four administrations (p = .867). These results indicate that playfulness meets both of the stability requirements for personality traits—stability across both time and situations.

Administration	N	Mean	StDev
Initial	60	21.767	9.039
Windows	62	22.339	8.248
Word	60	22.250	8.171
Excel	49	21.061	9.355

To further test the playfulness as stable trait hypothesis, a two-way ANOVA (see Table 5 below) was computed to determine whether it was by subject or by time that the results were changing. If the playfulness trait is dynamic, one would expect to see changes occur over time. However, if it is stable, one would expect to see the effect by subject. The participants themselves (SUBJECTS) accounted for the variance (F = 13.178, p = .000), while different administrations (TIME) did not have a significant effect (F = 1.300, p = .276). The variation in an individual’s results can be attributed to the individual’s playfulness trait, not the timing of the administration.

			  Sum of                Mean           Signif
Source of Variation      Squares     DF       Square       F     of F

Main Effects           14762.725     75      196.836    12.698   .000
  TIME                    60.469      3       20.156     1.300   .276
  SUBJECTS             14707.905     72      204.276    13.178   .000
Explained              14762.725     75      196.836    12.698   .000
Residual                2402.781    155       15.502
Total                  17165.506    230       74.633
Table 5: ANOVA Results

The results of this longitudinal study indicate that playfulness is a stable trait. The playfulness score is consistent, measures a single factor, and remains somewhat static. Moreover, means and standard deviations were stable over time. This study also supports the reliability of Webster & Martocchio’s (1992) operationalization of the playfulness construct. Their seven-item Computer Playfulness Scale demonstrated internal consistency, unidimensionality, and temporal and situational stability as evidenced by Cronbach’s alpha, factor validity, and test-retest correlations.

Previous researchers have suggested adapting training methods based on trainee characteristics (e.g., Bostrom, Olfman, & Sein, 1988; Bostrom, Olfman, & Sein, 1990; Wexley, 1984). Past studies of training methods have been inconclusive, and external effects of those methods on training effectiveness were posited to depend on other factors, including characteristic attributes of the trainees (Tannenbaum & Yukl, 1992). More research is needed to develop and understand training method adaptations that best utilize the stable trait nature of playfulness.

This research supports the temporal stability and situational consistency of the playfulness construct. The subjects of this study demonstrated a marked stability in the playfulness trait as they gained experience in their computing environment and knowledge of a representative spectrum of software applications.

The resulting stable trait characterization of the playfulness construct has important implications to both IS academics and researchers. Although prior research associates playfulness with increased learning and performance, our research suggests that the stability of the playfulness trait will make attempts to manipulate individual playfulness unlikely to succeed. We would suggest that the playfulness construct may best be accommodated by matching system and individual playfulness.

MIS designers or trainers who wish to utilize the playfulness trait should be able to do soby performing a one-time playfulness assessment rather than conducting longitudinal measures on individuals. This is good news both to practitioners who are trying to build effective systems and to researchers trying to further investigate the playfulness construct. In particular, it greatly simplifies playfulness experimental design as it renders individual playfulness traits stable rather than dynamic.

The research limitations include those traditionally acknowledged in conjunction with the use of student subjects. More to the point, this research investigates the stability of the playfulness trait in a training environment of intermediate duration (5 weeks) and varying software to test stability across situations. Further investigation is warranted into the stability of the playfulness trait across longer periods (as might be encountered by end-users of systems) and across alternative situations. For instance, training type and style could influence individual playfulness.

Further research should build on the stability of the playfulness trait by examining the outcomes of manipulating playfulness in training. For instance, these authors are currently using treatments differing by playfulness items to investigate the interaction between individuals’ playfulness traits and the playfulness of the computing environment in determining outcomes such as training satisfaction, user satisfaction, and individual performance measures.

Further research should be conducted into mechanisms by which playfulness enhances training or system performance. The proper matching of system and user playfulness to manipulate user mood offers one interesting avenue of research. A continuous stream of research has associated mild mood elevation with enhanced creative thinking (e.g., Richards, 1993; Eckblad & Chapman, 1986; Schuldberg, 1990 ), improved problem solving (Greene & Noice,1988), and better comprehension of new concepts (Jamison, 1989). Proper matching of system and/or training playfulness with individual playfulness characteristics may offer an opportunity to manipulate user mood with a potential outcome of better system performance.

Alreck, P. L., and Settle, R. B. (1995). The Survey Research Handbook (2nd edition), Chicago: Irwin Professional Publishing.

Anastasi, A. (1988). Psychological Testing (6th edition), New York: Macmillan Publishing Company.

Barnett, L. A. (1990). "Playfulness: Definition, Design, and Measurement," Play and Culture, Vol. 3, 319-336.

Barnett, L. A. (1991). "The Playful Child: Measurement of a Disposition to Play," Play and Culture, Vol. 4, 51-74.

Blau, G. J. (1988). "Further Exploring the Meaning and Measurement of Career Commitment," Journal of Vocational Behavior, Vol. 32, 284-297.

Bostrom, R. P., Olfman, L., and Sein, M. K. (1988). "End-User Computing: A Research Framework for Investigating the Training/Learning Process," Human Factors in Management Information Systems, Norwood, NJ: Ablex Publishing Corporation.

Bostrom, R. P., Olfman, L., and Sein, M. K. (1990). "The Importance of Learning Style in End-User Training," MIS Quarterly Vol. 14, No. 1, 101-119.

Buss, A. H. (1989). "Personality as Traits," American Psychologist, Vol. 44, 1378-1388.

Compeau, D. R., and Higgins, C. A. (1995). "Computer Self-Efficacy: Development of a Measure and Initial Test," MIS Quarterly, Vol. 19, No. 2, 189-211.

Crano, W. D., and Brewer, M. B. (1973). Principles of Research in Social Psychology, New York: McGraw-Hill.

Cronbach, L. J. (1951). "Coefficient Alpha and the Internal Structure of Tests," Psychometrika, Vol. 16, 297-334.

Csikszentmihalyi, M. (1975). Beyond Boredom and Anxiety, San Francisco: Jossey-Bass.

Eckblad, M., and Chapman, L. J. (1986). "Development and Validation of a Scale for Hypomanic Personality," Journal of Abnormal Psychology, Vol. 3, 214-222.

Fajou, S. (1996). "Computer Anxiety," <http://www.edfac.usyd.edu.au/projects/comped/Fajou.html>

Greene, T. R., and Noice, H. (1988). "Influence of Positive Affect Upon Creative Thinking and Problem Solving in Children," Psychological Reports, Vol. 63, 895-898.

Jamison, K. R. (1989). "Mood Disorders and Patterns or Creativity in British Writers and Artists," Psychiatry, Vol. 52, 125-134.

Jones, M.C., and Berry, R. L. (1995). "Information Technology: An Assessment of Student Perceptions," Journal of Computer Information Systems, Summer, 28-32.

Kappelman, L. A. (1995). "Measuring User Involvement: A Diffusion of Innovation Perspective," The DATA BASE for Advances in Information Systems, Vol. 26, No. 2 and 3, 65-83.

Litwin, M. S. (1995). How to Measure Survey Reliability and Validity, Vol. 7, The Survey Kit, Thousand Oaks, CA: Sage Publications.

Martocchio, J. J., and Webster, J. (1992). "Effects of Feedback and Cognitive Playfulness on Performance in Microcomputer Software Training," Personnel Psychology, Vol. 45, No. 2, 553-578.

Mason, R. O., and Mitroff, I. I. (1973). "A Program for Research on Management Information Systems," Management Science, Vol. 19, No. 5, 475-487.

McCrae, R. R., and John, O. P. (1992). "An Introduction to the Five Factor Model and Its Applications," Journal of Personality, Vol. 60, 175-215.

Miller, S. (1973). "Ends, Means, and Galumphing: Some Leitmotifs of Play," American Anthropologist, Vol. 75, 87-98.

Nash, J. E. (1990). "Working at and Working: Computer Fritters," Journal of Contemporary Ethnography, Vol. 19, No. 2, 201-225.

Nunnally, J. (1978). Psychometric Theory (2nd edition), New York: McGraw-Hill.

Richards, R. (1993). "Seeing Beyond: Issues of Creative Awareness and Social Responsibility," Creativity Research Journal, Vol. 6, 165-183.

Sandelands, L. E. (1988). "Effects of Work and Play Signals on Task Evaluation," Journal of Applied Social Psychology, Vol. 18, No. 12, 1032-1048.

Schuldberg, D. (1990). "Schizotypal and Hypomanic Traits, Creativity, and Psychological Health," Creativity Research Journal Vol. 3, 218-230.

Starbuck, W. H., and Webster, J. (1991). "When is Play Productive?" Accounting, Management, and Information Technology, Vol. 1, No. 1, 71-90.

Stewart, G. L., Carson, K. P., and Cardy, R. L. (1996). "The Joint Effects of Conscientiousness and Self-Leadership Training on Employee Self-Directed Behavior in a Service Setting," Personnel Psychology, Vol. 49, No. 1, 143-164.

Straub, D. W. (1989). "Validating Measurements in MIS Research," MIS Quarterly, Vol. 13, No. 2, 147-169.

Tannenbaum, S. I., and Yukl, G. (1992). "Training and Development in Work Organizations," Annual Review of Psychology, Vol. 43, 399-441.

Veenhoven, R. (1994). "Is Happiness a Trait? Tests of the Theory That a Better Society Does Not Make People Any Happier," Social Indicators Research, Vol. 32, No. 2, 101-160.

Webster, J., and Martocchio, J. J. (1992). "Microcomputer Playfulness: Development of a Measure With Workplace Implications," MIS Quarterly, Vol. 16, No. 2, 201-226.

Webster, J., Trevino, L. K., and Ryan, L. (1993). "The Dimensionality and Correlates of Flow in Human-Computer Interactions," Computers in Human Behavior, Vol. 9, 411-426.

Wexley, K. N. (1984). "Personnel Training," Annual Review of Psychology, Vol. 35, 519-551.

Widiger, T. A. (1993). "The DSM-III-R Categorical Personality Disorder Diagnoses: A Critique and an Alternative," Psychological Inquiry, Vol. 4, 75-90.

The following questions ask you how you would characterize yourself when using microcomputers. For each adjective below, please circle the number on the answer sheet that best matches a description of yourself when you interact with a microcomputer.