This chapter provides details of the methodological approach used in this study. It presents the research framework, the hypothesis to be tested and the research procedure. It also provides details on the questionnaires used in the study. The sampling procedure, data collection and data analysis process are also explained.
Research Framework
Figure 3.1 depicts the research framework used in the study. The demographic features, the two dimensions of the situational factor as well as the work values that will be examined are all outlined.
Figure 3.1. Research Framework
The counterproductive work behaviors which fall under misuse of resources are as follows: spend time on the internet for reasons not related to work, make personal calls at work, make personal copies at work, use email for personal purposes and play computer games during work time. Behaviors under misuse of information are as follows: discuss
Demographics: sex, Religious commitment, Work experience,
Socioeconomic status H1a –H1d
H2a-H2b Misuse of information
H3a-H3e
42
clients’ confidential matters with unauthorized personnel, lie to supervisors to cover up a mistake and intentionally fail to give coworkers necessary information (Gruys and Sackett, 2003).
Hypotheses
Hypothesis 1a: Males will generally display a higher propensity to engage in both misuse of resources and misuse of information.
Hypothesis 1b: Employees with high religious commitment will have a lower propensity to engage in both misuse of resources and misuse of information.
Hypothesis 1c: Work experience in negatively related to the propensity to engage in both misuse of resources and misuse of information.
Hypothesis 1d: Employees of high socioeconomic status will have a lower propensity to engage in both misuse of resources and misuse of information.
Hypothesis 2a: Consistency in CWB group norms will be positively related to the propensity of an individual to engage in both misuse of resources and misuse of information.
Hypothesis 2b: Amount of CWB group norms will be positively related to the propensity of an individual to engage in both misuse of resources and misuse of information.
Hypothesis 3a: A high value for economic returns is positively related to the propensity of an individual to engage in both misuse of resources and misuse of information.
Hypothesis 3b: A high value for independence is negatively related to the propensity of an individual to engage in both misuse of resources and misuse of information.
Hypothesis 3c: A high value for associates is positively related to the propensity of an individual to engage in both misuse of resources and misuse of information.
Hypothesis 3d: A high value for security is negatively related to the propensity of an individual to engage in both misuse of resources and misuse of information.
43
Hypothesis 3e: A high value for achievement is negatively related to the propensity of an individual to engage in both misuse of resources and misuse of information.
Hypothesis 4a: A high value for economic returns when moderated by group norms is positively related to the propensity to engage in both misuse of resources and misuse of information.
Hypothesis 4b: A high value for independence when moderated by group norms is negatively related to the propensity to engage in both misuse of resources and misuse of information.
Hypothesis 4c: A high value for security when moderated by group norms is negatively related to the propensity to engage in both misuse of resources and misuse of information.
Hypothesis 4d: A high value for associates when moderated by group norms is positively related to the propensity to engage in both misuse of resources and misuse of information.
Hypothesis 4e: A high value for achievement when moderated by group norms is negatively related to the propensity to engage in both misuse of resources and misuse of information.
Research Method
This study can be classified as a quantitative vignette study. Simply put this research combines a vignette experiment and a traditional survey. This method has also been referred to as a factorial survey. This is because it combines the ideas from multivariate experimental designs with sample survey procedures. A vignette is “a short, carefully constructed description of a person, object or situation, representing a systematic combination of characteristics” (Atzmuller & Steiner, 2010, p.128). This method has been used in the fields of psychology, sociology, marketing as well as education and training and is currently
44
gaining popularity in other fields such as Social Sciences. Atzumuller & Steiner (2010) describe it as “promising but too infrequently used” for investigating the beliefs, attitudes and judgments of respondents (p.128).
Vignettes allow a researcher to introduce and manipulate factors. The number of vignettes constructed represents a full factorial combination of all the factors under investigation (Taylor, 2006). One of the most unique aspects of a vignette is that it makes it possible to analyze the effects of people’s judgments through the systematic varying of the characteristics presented in the situation description (Alexander & Becker, 1978). As a result, it is especially useful for examining beliefs and decision making. The size of vignette samples used in research varies from small designs with only four (representing two factors with only two factor levels) to large designs with thousands. Usually respondents are not given an entire set of vignettes but a carefully selected subset (Rossi, Sampson, Bose, Jasso
& Passel, 1974). This study utilizes only four vignettes representing two (2) factors applied to two situations. Each respondent is given a subset of two. The procedure for constructing and assembling vignette packs outlined by Jasso (2006) as well as the factorial survey process outlined by Taylor (2006) was used to guide the methodology decisions of this stud y.
Vignettes are also an effective way to overcome the problems associated with self report (Alexander & Becker, 1978) especially when soliciting responses for topics (such as counterproductive work behaviors) that are considered socially undesirable/unacceptable.
Respondents may be unwilling to admit that they have engaged in CWB. However, when presented with hypothetical situations featured in vignettes, they may be more willing to concede and admit to the fact that they would engage in CWB. The fact that the vignette attempts to present a realistic situation that may actually exist in the work environment of respondents (Veal, 2002) makes it possible for this research to accurately capture
45
respondents’ propensity to engage in the listed CWBs.
In addition, vignettes allow for flexibility in research design as they can be presented in different forms such as text vignettes (in keyword, dialogue or narrative style) or as cartoons, pictures, audio or video. The form adopted in this study is the narrative style. For all of the reasons presented in the preceding paragraphs, the choice of methodology to be used in this study is valid and justifiable.
Sample Setting
The population of interest in this research is graduates of the Business Administration department of the Sir Arthur Lewis Community College who are between the age of 19 - 25 and who have had at least three months of full-time or part-time work experience. The total population is approximately 3000 – 4000 students (calculations based on statistics from www.stats.gov.lc). This population was chosen because this is the largest group within the population (stats.gov.lc) and graduates of Business Administration occupy a large number of positions within the public and private sector.
Instruments
The instruments consisted of four sections. Section one solicited demographic information, section two contained questions on work values, section three presented the vignette sets and questions on counterproductive work behavior and section four contained the social desirability scale.
Work values
The section on work values was adapted from Super’s Work Values Inventory.
Permission was sort from Donald E. Super’s son who holds the rights to Super’s Work Values Inventory. This is a scale of 45 items designed to test 15 work values. Items relevant to the work values to be examined were extracted; each work value is represented by three items. The wording of three statements was changed to ensure respondents’ comprehension.
46
Statement 8 was changed from “are one of the gang” to “are part of the team or crew”. The word “crew” is a colloquial term commonly used by the population from which the sample will be drawn. Statement 34 was changed from “have good contacts with fellow workers” to
“have frequent contact with fellow workers”. Statement 39 was changed from “are paid enough to live right” to “are paid enough to live well”. The following are sample items from the work values section of the questionnaire: “can get a raise” for economic returns, “have freedom in your own area” for independence, “have frequent contact with fellow workers”
for associates, “are sure of always having a job” for security, “get the feeling of having done a good days work” for achievement. Please refer to the appendix for remaining items.
Group norms and counterproductive work behavior
Four vignettes were developed for this study. The first two presented a situation that dealt with misuse of resources. The first two vignettes had the following features: (a) length of tenure (three months) (b) opportunity (equal access to resources) (c) work group size (6 persons of undefined gender) (d) one feature of group behavioral norms (amount or consistency) (e) description of observed counterproductive work behavior (misuse of resources). The second two dealt with misuse of information and had the following features:
(a) opportunity (equal access to information) (b) one feature of group behavior norms (amount or consistency) and (c) description of observed counterproductive behavior (misuse of information).
The measure for counterproductive work behavior was based on Gruys and Sackett (2003). Four items depicting misuse of resources were chosen from the category labeled misuse of time and resources and three items were chosen from the misuse of information category. The items were chosen based on the researcher’s interest and also the ease with which the stated behavior could be captured in a vignette. Sample items for misuse of resources are as follows: “spend time on the internet for reasons not related to work” and
47
“make personal calls at work”. Sample items for misuse of information are as follows: “talk to other people about clients” and “retain information for your own use when you are supposed to pass it on to a coworker” (refer to appendix for remaining items).
Social desirability
The social desirability measure was taken from Strahan and Gerbasi (1972). It is a short version of the Marlow-Crowne Social Desirability Scale consisting of ten (10) true or false items. Sample items for Social desirability are as follows: “I am always willing to admit when I make a mistake”, “I have never deliberately said something that hurt someone’s feelings” and “there have been occasions when I felt like smashing things”.
Three separate questionnaires (A, B and C) were complied. All three had identical content for demographics, work values, counterproductive behavior and Social Desirability.
Questionnaire A had a vignette for misuse of resources which portrayed amount in group behavioral norms and a vignette for misuse of information which also portrayed amount in group behavioral norms. Questionnaire B had a vignette for misuse of resources which portrayed consistency in group behavioral norms and a vignette for misuse of information which also portrayed consistency in group behavioral norms. Questionnaire C portrayed no group norms.
Before they were posted online, the questionnaires were piloted with a group of 30 respondents (St. Lucian students studying in Taiwan). Respondents were asked to provide feedback about the quality of the vignettes and the clarity of the language used in the questionnaire. This feedback resulted in several modifications being made to the wording of the questions in section three.
Reliability testing of the instrument after data collection produced the results reported in table 3.1. The cronbach alpha for each of the scale variables is reported and as shown in the table is considered to be acceptable. Originally, each work value was measured by three
48
items, however after they were tested for reliability, four questions were deleted to make the measure for the individual work values more reliable.
Table 3.1
Results of Reliability Testing for Scale Items
Variable Cronbach Alpha No. of Items
Religious commitment .791 2
Economic Returns .759 2
Independence .674 2
Associates .739 2
Security .788 2
Achievement .634 3
Misuse of Resources .790 5
Misuse of Information .700 3
Data Collection
Questionnaires were posted online (using Google docs) and Facebook as well as other social networking sites was used to reach suitable participants. This was the best way to reach this population because they are widely dispersed. Moreover this age group is technologically savvy and actively participates in online networking. For these reasons, a platform which utilizes technology was the best means to contact and engage them. Three questionnaires were posted, each presenting a different vignette set. The first questionnaire (A) will contained two vignettes, one presenting amount in group behavioral norms for misuse of information and the other presenting amount in misuse of information. The second questionnaire (B) also contained two vignettes, one presenting consistency in misuse of resources and the other presenting consistency in misuse of information. The third questionnaire (C) presented two vignettes stating that there are no observable group norms.
Links were randomly sent to participants ensuring that the same number was sent for each of the questionnaires.
49
Approximately 300 participants were contacted and invited to complete the questionnaires for this research. Data collection was conducted during the months of February and March. First a message, inviting recipients to participate in the research, accompanied by the questionnaire link was sent. This was followed (a week later) by a reminder. The researcher also used Facebook instant messaging to encourage participation.
A total of 203 questionnaires (Questionnaire A = 67 responses, Questionnaire B = 68 responses and Questionnaire C = 68 responses) were collected, resulting in a response rate of approximately 67%.
This study utilized the concepts of experimental design. The following visually represents the method that was used to collect data from participants.
Section B of Questionnaire Random Selection
Figure 3.2. Data collection process Participants N=203
Section A of Questionnaire Demographics and Work values
(C) Control Group Group Behavioral Norms (A)
Amount (MoR), Amount (MoI)
Group Behavioral Norms (B) Consistency(MoR),Consistency (MoI)
Questions on Counterproductive work behavior
Social Desirability Measure
50
Data Analysis
SPSS version 19 was used to conduct statistical analysis of the data collected for this study. After data was collected, descriptive statistics (mean, standard deviation, frequency) and Pearson’s correlation was used to give an overview of the sample profile as well as an overview of the relationship between variables.
Hypothesis testing was then carried out using a variety of methods. Hierarchical regression was be used to test hypotheses 3a to 3d and 4a -4e. This is because these measure the relationship between counterproductive work behavior and independent variables that are measured using a liker scale (work values and religious commitment). For testing the moderation effect featured in hypotheses 4a -4e, a new independent variable was created by first centering the continuous variable data (work values) and then multiplying the centered data by the categorical variable (group norms). Independent sample T-test was used to test hypotheses 1a. This is because the independent variables featured in this hypothesis are categorical variables with two categories (sex). ANOVA was used to test hypotheses 1c, 1d as well as 2a and 2b, since they feature categorical independent variables represented by more than two categories (work experience, socioeconomic status and group norms including the control group).
51
Research Procedure
Figure 3.3 outlines the research procedure followed during the course of this study.
Figure 3.3. Research procedure Review literature
Establish groundwork for the study:
topic, methodology, framework, hypotheses
Make decisions about sample, sampling procedure
Develop the instrument: Get permission to use measures where necessary
Expert review and pilot test instrument: make modifications where necessary
Data collection Data analysis
Present findings, recommendations and conclusions
52
53