Student Leader Representation Networks

Get Complete Project Material File(s) Now! »

Stimulus Development 


In order to develop the Video Vignette Behavior Selection Task (VVBST), which is the main stimulus of the focal study, this pilot study sought to establish the extent to which a series of behavioral statements represent traits commonly used to describe leaders.  In order to collect the information, 30 Virginia Tech students (21 female), aged 18 to 22 (avg. = 18.7), were recruited via the SONA system at Virginia Tech and participated in the pilot study in exchange for extra credit in a psychology course.  In order to achieve a participant pool that generally represented a university student sample, all participants were required to be 18 years of age. Additional criteria for participation were access to the internet via a computer and that the study was completed during one sitting.


Participants completed an online survey, developed using Snap Surveys’ Snap 8 survey software.  The survey was based on a list of discrete behaviors representative of ILT trait factors suggested by Offerman, Kennedy, & Wirtz (1994), and Epitropaki & Martin (2004).  The behavioral sentences were derived from a variety of sources that addressed leadership behaviors and those associated with the factors described by Epitropaki & Martin (2004).  (see Appendix A for a trait-behavioral statement list). Survey presentation required participants to make responses over the span of 10 survey pages.  The first survey page collected general demographic and consent information.  Pages two through nine asked the participants to read a series of statements and then rate the statement on how well it described the behavior of a person represented by the underlined word, such as “A person who is intelligent can think quickly when questioned.”  Seven trait-behavior statements were presented on each page, beginning on page two, through to page nine.  The traits and behaviors were paired so that throughout the survey, each trait was listed with two behavioral statements meant to reflect the trait, and two behavioral statements that were meant to reflect a different trait.  The traits or behaviors were not presented more than once on each survey page.  In addition, no trait or statement appeared as the last item on a page and then as the first item on the following page.  The final page of the survey collected general information on the understanding of Virginia Tech’s honor system, the use of single sanction systems, and support of a single sanction system at Virginia Tech .

Design and Procedure

Once registered for the study, participants were directed to an online survey.  The survey asked participants to provide consent and some general demographic information.  Upon providing consent, participants began the survey with instructions to read a series of behavioral statements and to rate the statements on “how well [the statement] describes the behavior of a person described by the underlined word.”  Participants rated 56 statements on a 7-point Likerttype scale, ranging from 1 (not at all representative) to 7 (extremely representative). Upon completion of the rating task, participants completed a series of questions about opinions and preferences with regard to the honor system at Virginia Tech.  Upon completion of the online survey, participants were thanked for their participation and directed to a debriefing page.


The rating data were analyzed in order to select the 10 trait-behavior statements for use in a second pilot study and the development of the video stimulus of the main study.  The means for the representative trait-behavior statements ranged from 3.47 to 6.53, while the means for the non-representative trait-behavior statements ranged from 1.43 to 6.27 (see Table 1 for means and t-test information).  Examinations of the ranges suggest that the trait-behavioral statements are not all perceived as accurately representing the related traits, and that there are differences across behavioral statements with regard to the representative traits.  In order to discover which traitbehavioral statements provided behaviors that were most representative of leadership traits, the data were analyzed by first conducting a one-sample t-test for each item (Scott & Brown, 2004).  While Scott & Brown used a test value of 4.0 to indicate a significant difference above neutral, a test value of 4.5 was used for the tests in order to arrive at trait-behavior statements that were well above neutral. The one-sample t-tests revealed that of the 56 trait-behavior statements that were rated, 36 of the statements were significantly different from 4.5, at the 0.05 level.  Of the 36 significantly different statements, 21 were significant and in the desired direction, the rating range of these statements was 5.27 to 6.53.  The next analyses determined if each behavioral statement was perceived differently across the two traits each statement was paired with.  For example, is there a significant difference between the rating of the statement of “Exhibits an openness to the ideas of other people” when paired with the trait of “caring” or the trait “charismatic?”  Paired t-test revealed that 20 of the 21 pairs were significantly different from one another at the 0.05 level.  Of the 20 significantly different pairs, 18 were in the desired direction where the intended trait was rated significantly higher than the non-intended trait.  Fifteen of the behavioral statements, representing 10 traits met the requirements.  Within the traits with two behavioral statements meeting the criteria mentioned above, the statement with the largest mean difference on the paired-sample t-test was selected to represent the trait.  The resulting 10 traits and related behavioral statements are presented in Appendix C. Of the traits selected, one-way ANOVA revealed two trait-behavior ratings that were rated differently as a function of participant sex.  The different ratings were for the behavioral statements associated with honest and motivated.  Another series of one-sample t-tests were conducted on the data provided by male and female participants.  For female participants, a significant difference greater than 4.5 was found for ratings of the motivated (t=8.96, p=.000), and honest (t=6.79, p=.000) behavioral statements. For male participants, a significant difference greater than 4.5 was not found for ratings of the motivated (t=1.386, p=.203) and honest (t=.559, p=.592) behavioral statements.  While the male difference is not desirable, the number of male participants is small and may account for the inability to find a statistically significant difference in ratings.   Based on the small number of male participants, and the fact that the ratings were still above the midpoint, the two trait-behavior statements were not excluded from the study. Analyses of the items relating to the honor system found that on average, the students were familiar with the honor system in place at Virginia Tech (avg.=5.23, SD=1.14).  With regard to the single sanction honor system ratings (avg.=3.67, SD=2.07), the ratings were significantly lower when compared to familiarity with the Virginia Tech system (t=-4.379, p=.000).  Generally, while not as familiar to the participants, the single sanction system was not supported by the students (avg.=3.1, SD=1.81).  The ratings were not significantly different as a function of participant sex.


The trait-behavior pairs were presented in such a fashion that those trait-behavior pairs that were not the intended trait-behavior pairs were closely related to the intended trait-behavior pair.  The decision to have highly related traits and behaviors linked to one another was based on the desire to determine which behaviors most saliently represent the traits.  The 10 trait-behavior pairs that were selected seem to discriminate between traits and behaviors, as well as having behaviors that adequately represent the intended traits. The issue of selecting a subject matter that was relevant to the student sample was less clear from the results.  The average rating of the students with regard to the implementation of the single sanction system, and familiarity with the single sanction system indicate a relatively neutral position on the issue of the single sanction system.  Because of pilot one’s results, pilot study two explored the alternative topic of file sharing, as the issue may be more relevant to the student sample, and is one that is being actively addressed in universities across the nation and by the Recording Industry Association of America.

Stimulus Development

The main manipulation (VVBST) relied heavily on the observation of the behavior of potential leaders, in order to induce a participant to select a leader exhibiting behaviors that were cognitively misaligned or cognitively aligned, depending on the topic relevance to the participant.  While pilot study one provided a foundation for understanding which behavioral statements represent leadership traits, pilot one does not provide information on how traits and/or behaviors are structured cognitively.  In order to assess how the behaviors exhibited by leaders are related to the traits understood to represent leaders, the study described below provided insight on the trait and behavior linking between mental models, and provided a means of developing the prototypes for leaders exhibiting behavioral patterns that were either aligned with cognitive expectations, or misaligned with cognitive expectations.  The study also assessed the relevance of file-sharing issues to the student sample used across all of the studies.


Sixty-nine undergraduate students (50 female, 19 male, avg. age = 19.52) participated in the study in exchange for extra credit in an undergraduate course. In order to achieve a participant pool that generally represented a university student sample, participants were at least 18 years of age and enrolled full-time in a four-year university during the regular academic year.


Participants performed a series of concept-rating tasks.  The rating tasks were presented to participants via Dell workstations present in University computer labs.  The rating tasks were administered using E-prime (Psychology Software Tools).  There were two distinct rating tasks.  One of the concept-rating tasks asked participants to rate the relatedness of n(n-1)/2 word pairs, based on the ten traits selected from the first pilot study, for a total of 45 trait pair ratings. The other concept-rating tasks asked participants to rate the relatedness of n(n-1)/2 statement pairs, based on the ten behavioral statements selected from the first pilot study, for a total of 45 statement pair ratings (See appendix C for traits and behaviors). All of the ratings were placed in the context of thinking about how the concept pairs are related within the context of a student leader’s behaviors or traits, and used the same scale as pilot. During the rating tasks, participants were presented with each pair rating, with the words/sentences centered horizontally on the screen.  One of the pair items was presented in the top position, while the other was presented in the bottom position.  For example, on the computer monitor, a participant rating the pair of “caring” and “charismatic” was presented with a statement about the rating context and task at the top of the screen, followed by a presentation of the rating scale.  Beneath the statement and scale, participants were presented with the word “caring” horizontally centered on the screen and slightly above the center of the screen, while the word “charismatic” was centered horizontally and slightly below the center of the screen. Participants used the keyboard to enter their numeric rating of the pair, at which point, the program presented the next pair to be rated. The rating task continued until all of the pairs in a concept-rating task were complete. There was not a time limit on the concept rating task, yet a response time was recorded along with the responses. Pair presentation was randomized within each concept-rating task, and the order of the two concept-rating tasks were counter-balanced so that half of the participants performed the concept-rating task related to traits first, and the other half performed the concept-rating task related to the behavioral statements first.   In addition, if the trait term for “caring” was presented in the top position above “charismatic,” the behavioral statement linked to “caring” was presented in the bottom position when presented with the behavioral statement related to “charismatic.” The cognitive task consisted of a word search exercise, which served as a distracter task between the two rating exercises.  The word search was developed using an online word search tool and contained twenty words selected from  The twenty words were selected based on the belief that they are words that are not used in everyday language and would not be highly familiar to the participants. The word list and the word search are presented .

1.1 First-order theories of leadership
2.1 Stimulus Development
2.2 Stimulus Development
2.3 Stimulus Development
3.1 Participants
3.2 Student Leader Representation Networks
3.3 Topic Relevance Measure/File Sharing Attitude
3.4 Manipulation Checks
4.1 Leadership and Connectionism Findings
4.2 Future Research
Understanding leader representations: Beyond implicit leadership theory

Related Posts