Back to Table of Contents

 

 

 

CHAPTER 6

RESEARCH DESIGN AND PROCEDURES

Introduction

The underlying purpose of this study is to explore the concept of critical worldview integration and the volitional processes involved.   Specifically, are there any identifiable epistemological or religious beliefs that appear to be conducive to integration of this sort?   Additionally, are there any specific beliefs concerning the learning process or motivational factors that are present (or absent) among those who engage in critical and volitional integration of their worldview?   If so, is evidence of these beliefs or characteristics helpful in either predicting a person’s degree of engagement or in encouraging an increase in such activity?  

Due to the difficulty of directly observing and measuring introspective processes and structures such as worldview integration, this study will attempt to investigate the concept through its relationship with a few domains that may be more easily quantified.   These domains consist of: a) two categories of worldview belief—religious worldview (Chapter 2) and epistemological worldview (Chapter 4); b) beliefs concerning the learning process (Chapter 4); and c) two psychological constructs—motivation to learn (Chapter 5) and subjective well-being (Chapter 5).   Motivation to learn is further limited to intrinsic and extrinsic motivation within a self-determination theory of motivation.   Subjective well-being is divided into two sub-domains; satisfaction with life and subjective vitality.   In addition, the domain of epistemological worldview is considered by many to be its own psychological construct within the burgeoning field of personal epistemology.   Previously developed instruments will be used to quantitatively measure each of these domains.   Finally, some analysis will be done regarding the family upbringing and personal belief questions posed at the end of the survey.

 

Research Questions

Although the overall research question is to determine if there are specific characteristics that are present (or absent) in those who engage in critical and volitional integration of their worldview, it is helpful to divide this project into four specific questions that allow for more operationalized evaluation.  

            Question 1.       What differences are there between the students’s   gender, their religious upbringing or their willingness to seek independent answers and their scores on measures of religious worldview, personal epistemology, intrinsic motivation, personal vitality, and satisfaction with life?   (This question treats religious worldview as an interval measure).

             Question 2.       What differences are there between the upper level samples and the lower level sample concerning their scores on measures of religious worldview, personal epistemology, intrinsic motivation, personal vitality, and satisfaction with life?   (This question treats religious worldview as an interval measure).

             Question 3.       What differences are there between the different religious worldview categories (as a categorical score) and their scores on measures of personal epistemology, intrinsic motivation, personal vitality and satisfaction with life?  

             Question 4.       What correlations can we discover within each sample and the combined participants regarding personal epistemology, intrinsic motivation, personal vitality, satisfaction with life, and one’s religious worldview?   Do these correlations suggest any common factors that might be anticipated to promote a willingness to engage in worldview integration?   (This question treats religious worldview as an interval measure).

 

Within each of these questions, there are a number of subsidiary questions that may be present and investigated.   Since this research is generally inductive in nature, it is important to set forth a series of null hypotheses encompassed within each of these subsidiary questions. Each is numbered (1.1 to 4.6) based upon which of the four research question (listed above) it pertains to.

 

Main effects of Gender:

            H 0 1.1:              There is no significant difference between gender in any of the five measured dimensions on the Epistemic Beliefs Inventory or the seven types of motivation found on the Academic Motivation Scale.

            H 0 1.2:              There is no significant difference between gender on either the Satisfaction with Life Scale or the Subjective Vitality Scale.

            H 0 1.3:              There is no significant difference between gender on religious worldview scores.

 

Main effect of family religious upbringing:

            H 0 1.4:              There is no difference between those students with parents who placed a strong emphasis on religious beliefs and those with parents who did not regarding any of the five measured dimensions on the Epistemic Beliefs Inventory or the seven types of motivation found on the Academic Motivation Scale.

            H 0 1.5   :            There is no difference between those students with parents who placed a strong emphasis on religious beliefs and those with parents who did not on their scores on either the Satisfaction with Life Scale or the Subjective Vitality Scale.

            H 0 1.6:              There is no difference between those students with parents who placed a strong emphasis on religious beliefs and those with parents who did not regarding their religious worldview scores.

 

Main effect of seeking independent answers:

            H 0 1.7:              There is no difference between those students who have indicated that they have grappled with investigating their own religious beliefs independent of their parents and those who have not regarding any of the five measured dimensions on the Epistemic Beliefs Inventory or the seven types of motivation found on the Academic Motivation Scale.

            H 0 1.8:              There is no difference between those students who have indicated that they have grappled with investigating their own religious beliefs independent of their parents and those who have not regarding their scores on either the Satisfaction with Life Scale or the Subjective Vitality Scale.

            H 0 1.9:              There is no difference between those students who have indicated that they have grappled with investigating their own religious beliefs independent of their parents and those who have not regarding their religious worldview scores.

 

Main effect of educational level:

            H 0 2.1:              There is no difference between lower level students and upper level students regarding any of the five measured dimensions on the Epistemic Beliefs Inventory or the seven types of motivation found on the Academic Motivation Scale.

            H 0 2.2:              There is no difference between lower level students and upper level students regarding their scores on either the Satisfaction with Life Scale or the Subjective Vitality Scale.

            H 0 2.3:              There is no difference between lower level students and upper level students regarding their religious worldview scores.

 

Religious Worldview:

            H 0 3.1   :            There is no significant difference between those falling in different religious worldview categories regarding any of the five measured dimensions on the Epistemic Beliefs Inventory or the seven types of motivation found on the Academic Motivation Scale.

            H 0 3.2   :            There is no significant difference between those falling in different religious worldview categories regarding their scores on either the Satisfaction with Life Scale or the Subjective Vitality Scale.

            H 0 4.1:              There are no significant correlations between any of the measured factors on the Academic Motivation Scale and one’s score on the Religious World View Scale.

            H 0 4.2:              There is no significant correlation between any of the five measured dimensions on the Epistemic Beliefs Inventory and one’s score on the Religious World View Scale.

            H 0 4.3:              There is no significant correlation between scores on either the Satisfaction with Life Scale or the Subjective Vitality Scale and one’s score on the Religious World View Scale.

            H 0 4.4:              There is no significant correlation between any of the five measured dimensions on the Epistemic Beliefs Inventory and the seven types of motivation found on the Academic Motivation Scale.

            H 0 4.5:              There is no significant correlation between any of the five measured dimensions on the Epistemic Beliefs Inventory and one’s scores on either the Satisfaction with Life Scale or the Subjective Vitality Scale

            H 0 4.6:              There is no significant correlation between any of the seven types of motivation found on the Academic Motivation Scale and one’s score on either the Satisfaction with Life Scale or the Subjective Vitality Scale.

 

Research Design

Due to the exploratory nature of this study, a quantitative methodology has been selected.   At this stage, we have well documented research suggesting the domains that would be expected to play a role.   It is now important to investigate these suspected relationships and differences further in an attempt to provide more validity for their existence and to possibly gain more insight into the degree and direction of their interaction.   Later, as these relationships become better validated (or should it become apparent that these domains are not helpful), it may become necessary to return to a more in-depth qualitative observation of worldview integration and the factors involved at an individual or case level.  

In addition, a cross-sectional methodology will be used.  The purpose of this study is only to measure what currently is and not to see how a person’s willingness to integrate, their style of motivation, or their epistemic and religious worldview change over time.   A longitudinal study may be helpful at a later date to see whether changes in these domains result from specific forms of intervention—for example, one professor wondered what effect his freshman class entitled “Foundations of Christian Thought” has on his students over time—however this is not the purpose of this study.

 

Survey Instrument

In order to provide for quantitative and operational measurement, this study will use a questionnaire constructed from a series of instruments designed to measure the domains of religious worldview beliefs, motivation (from a self-determination theory), personal epistemological beliefs, and subjective wellbeing (both satisfaction with life and subjective vitality).   The research review for each of the domains has previously investigated the available instruments and the research behind them.

The instrument chosen to measure religious worldview beliefs is the most limited.   The Religious World View Scale (McLean, 1952; Jennings, 1972) measures the degree of theistic beliefs along a traditional Christian continuum.   Although data concerning validity and reliability is lacking, Jennings (1972) showed some internal validity by establishing a high correlation between this scale and Hogge & Friedman’s (1967) Scriptural Literalism Scale.   Still, Jennings (1972) was able to provide some norm scores for men (79.7, SD=20.1) and for women (89.3, SD=19.3) attending a Dallas junior college.   While the scale will arguably distinguish between an atheist, a theist with pluralistic and liberal tendencies, and an orthodox Christian, it will likely yield little in the way of distinguishable differences between a theistic and a truly non-theistic form of religion.   It is likely that a Buddhist, for example, will measure low on the theism scale and be scored as either an atheist or a very liberal theist.   Likewise, an orthodox Jew or Muslim will likely register in a similar fashion due to their differences in belief with orthodox Christianity.

McLean’s initial scale used 25 items rated on a 1 to 4 Likert scale totaling 100 points.   He also provided a fifth choice of “uncertain” or “no opinion” for 0 points.  Due to problems noted in Chapter 2, the “uncertain / no opinion” choice was eliminated and the Likert scale was expanded to 6 points, thus continuing to eliminate a middle position and providing a bit more separation to provide some insulation against ceiling effects within the evangelical Christian sample.   Thus, the scale will provide a score from 25 to 150.   This study will use the scale initially as an interval (not ratio) scale.   However, due to questions about its validity as an interval scale, this study will also use the scores to separate participants into five ordinal categories [naturalistic humanism 1 – (25 – 50); humanism (51 to 75); religious liberalism (76 to 100); liberal Christian (101 to 125); and strong Christian orthodoxy (126 to 150) for separate analysis.

For the domain of personal epistemology, a number of tests exist.   While no one test has been universally accepted, this study has elected to use the Epistemic Beliefs Inventory (EBI) (Schraw, Bendixen & Dunkle, 1995).   Validity is suggested by a factor analysis of the EBI that established five clear factors and explained 64% of the sample variation compared with only 39% from its closest rival (Schommer EQ) (Schraw, Bendixen & Dunkle, 2002).   In addition, retest correlations after one month yielded scores for each factor ranging from .66 to .81. This was better than the EQ and suggests decent reliability.   This inventory consists of 32 questions that yield both a total score and a separate score for five different dimensions (factors) labeled as: a) fixed ability; b) certain knowledge; c) omniscient authority; d) simple knowledge; and e) quick learning.   In this study, it is anticipated that the factor of omniscient authority will be important and the EBI has been shown to be more effective in discerning this factor than has the best alternative instrument (Schommer’s EQ).

For the domain of motivation (using a self-determination theory), at least two good tests exist, however the object of one’s motivation must be taken into account (motivation to learn versus other types of motivation) and the context (specific class versus global).   As a result, this study will use the Academic Motivation Scale – college version (Vallerand, Pelletier, Blais, Brier, Senecal & Vallieres, 1992).  This instrument consists of 28 questions yielding a total score along with seven sub-scale scores (amotivation, external motivation, introjected motivation, identified regulation, intrinsic motivation toward knowledge, intrinsic motivation to accomplish or create, and intrinsic motivation to experience sensations).   Confirmatory factor analysis has supported the internal validity of the instrument’s seven subscales (Fairchild, et al., 2004), while others have found that the subscales appropriately correlated with similar instruments attempting to measure similar constructs (Cokely, et al., 2001).   While the other instrument that was investigated (the MSLQ) (Pintrich, et al., 1991) also has high validity and reliability data, that instrument is directed at investigating specific classes and disciplines rather than a global motivation toward learning.

Finally, for the domain of subjective well-being, there is only one definitive test available for each concept (satisfaction with life and subjective vitality).   This study will use both the five item Satisfaction with Life Scale (SWLS) (Diener, et al., 1985) and the six item Subjective Vitality Scale (SVS) (Ryan & Frederick, 1997).   Each scale utilizes a 7-point Likert scale ranging from “not at all true for me” to “very true for me” and yields only one total score for each scale.   Internal validity for both scales seems strong and is established both by factor analysis (Kasser & Ryan, 2001) and by high correlations with other previous measures of satisfaction and a number of related concepts (eg. self-esteem, self-actualization, self-determination, physical illness and depression).   In addition, studies have shown that both scales measure slightly different aspects of well-being and do not duplicate measures of happiness (Nix, et al., 1999).   Ryan and Frederick (1997) also found test/retest reliability at an acceptable .64.   Two other items have been created by the researcher to be included with these two instruments.   The participants will be asked to respond in the same manner (7-point Likert scale) to the statements: 1) I am energized by a sense of awe and wonder; and 2) I look forward to learning new things each day.   These two questions will not be included in the satisfaction with life score or the subjective vitality score, but may lend some additional information and understanding to the issue of worldview integration.

The study also uses several additional researcher generated questions (questions 72 – 73 and questions 104 - 120.   Question 72 asks to what extent the participant is “energized by a sense of awe and wonder.”   This arguably investigates a possible measure of subjective vitality that also implicates the idea of wonder, and possibly intrinsic motivation.   Question 73 asks whether or not the participant “looks forward to learning new things each day.”   This investigates another possible measure of satisfaction with life – a measure that also implicates the idea of intrinsic motivation to learn.

Questions 104 and 105 investigate family encouragement toward education, while question 107 investigates family encouragement toward religion.   Questions 106, 109 and 110 investigate the actual religious beliefs of the participant’s parents.   Do one or both of the parents have “strong religious beliefs,” do they believe that one’s religious beliefs are more important that education, and finally, do they expect that their children will embrace the same belief?   Question 110 investigates whether the participant’s beliefs are, in fact, similar to those of one or both parents, while question 108 asks whether the participant was required to attend religious services regularly as a child.   Finally, question 112 asks for a self-report concerning whether the participant believes that they “have thought about and decided upon my own religious beliefs independent of the beliefs of either of my parents.”   This question provides a rudimentary self-indication concerning whether or not that participant has actively engaged in integration of their religious worldview.

Questions 113 to 120 ask the participant to report on certain personal beliefs that are held and the stability of those beliefs.   For example, questions 114 and 115 ask whether the participant has recently considered changing religious beliefs – or if these beliefs are currently in crisis or undergoing change?   Finally, question 120 investigates the idea of the natural feeling of being pursued or sought after by God as anticipated by Chapter 3 and Romans 1.   These questions serve to provide a certain richness to the study and may allow for some indication of whether a person currently, or has previously, engaged in true worldview integration.

Permission has been granted to the researcher for the use of the Epistemic Beliefs Inventory  (Schraw, Bendixen & Dunkle); the Academic Motivation Scale – college version (Vallerand, et al.); the Satisfaction with Life Scale (Diener) and the Subjective Vitality Scale (Ryan).   Copies of the letters granting permission are attached as Appendix A.   The Religious World View Scale (McLain) is apparently in the public domain.   Both McLean (1952) and Jennings (1972) are deceased, the publication where the scale was originally published (Motive) is no longer in existence, and the Journal For the Scientific Study of Religion has indicated that they hold no rights to the scale.   Review of a number of instrument and measurement inventories, including Measures of Religiosity (Boivin, 1999) list no owner, copyright holder, or contact information.

 

Sample Selection

It was first necessary to estimate a sample size for the study.   Using an estimation of effect size (medium), a desired significance level of .05, and a desired power of .7, it was estimated that the required sample sizes should be in excess of 66 for each of the anticipated Pearson correlations, 100 for the independent t tests, and 126 for ANOVAs.   This estimate is based upon previous effect size estimates for the domains in question and the suggestions of Olejnik (1984) in his article entitled Planning educational research: Determining the necessary sample size.

This study will use four distinct samples collected using two different methodologies.   The four samples consist of a) freshmen and sophomore students at various random institutions in California; b) freshman and sophomore students engaged in an entry level college class at an evangelical Christian institution in California; c) students engaged in graduate studies at various random institutions in California; and d) students engaged in graduate studies (philosophy & education) at an evangelical Christian institution in California.   These samples will hopefully provide a broad range of students from a number of different religious worldviews, who represent a range of motivations for engaging in formal education, who portray a variety of personal epistemology beliefs, and who exhibit different degrees of vitality and satisfaction with life.   It is also hoped that this sample will include a variety of types of religious upbringing and degrees of willingness to seek their own independent answers (as opposed to those instilled by parents) to life’s questions.

The participants for the entry level students at an evangelical Christian institution in California consist of approximately 250 freshmen and sophomore students in a required class entitled Foundations of Christian Thought.   The first sampling of 268 students was taken in October of 2005.   These students had a period of 45 days in which to respond.   Here, 186 completed surveys were returned, 102 freshman, 63 sophomores, 19 juniors and 2 seniors.   However, this sample was already four weeks into a class that is designed to promote integrative critical thinking skills.   As a result, another 158 students from the same class were also tested in February 2006 during the first week of class to see if there is any mean difference based upon the four weeks of class already completed by the first sample.   Here, an additional 86 completed surveys were returned, 61 freshmen, 18 sophomores, and 7 juniors.   However, these students were only given 14 days to complete the survey.   At the time of the deadline, additional surveys were still coming in.   Thus, it is likely that more than 86 completed surveys would have been collected had the deadline been extended.   All junior and senior data were retained for later analysis, but were excluded from the analysis of freshmen and sophomore responses.

The participants for the graduate philosophy or education classes at an evangelical Christian institution in California consist of 80 students enrolled in the Masters of Philosophy and Ethics and the Masters of Christian Education programs at the same institution.   Here, 52 completed surveys were returned.

Obtaining participants from secular institutions in California proved to be more difficult.   Several schools that were contacted refused to participate.   Several others indicated that they would not permit participation unless the project was sponsored through their institution with appropriate credit to the appropriate department and faculty.   Such sponsorship would be difficult to obtain for dissertation work.   One institutional research director pointed out that sponsorship would be easier if it resulted in the receipt of grant funds to that institution.   Another simply insisted that California law would not permit data collection at a state school by an outside institution.   As an attorney, I was unable to find any such legislation or regulation.   Finally, one state school indicated that due to concern for student rights, research could only be done through a process that allowed the students to self-select the research they participated in from a web-site describing the purpose of each research project.   This process seemed to compromise any semblance of randomness.   Clearly, the sophomore effect is beginning to result in limited abilities to collect random data in California institutions.  

Based upon these difficulties and the fact that the survey was easily accessible online, the researcher elected to turn to e-mail solicitations.   While e-mail data was not available through the institutions, there are several opt-in e-mail list firms that maintain data filtering current students, their ages, and their year in school.   After contacting several such providers, Direct Web Advertising (www.directwebadv.com) was chosen and several e-mail campaigns were sent out about 1 month apart.   The first campaign consisted of Direct Web randomly selecting a mailing of 5,000 current undergraduate and graduate college students within California.   The e-mail mailing piece is attached as Appendix C.   From this mailing, completed surveys were received from 38 freshmen, 21 sophomores, 29 juniors, 15 seniors, and 101 graduate students.   Due to the imbalance of graduate students, the second campaign was limited to freshmen and sophomores with an additional e-mail filter of students from 18 to 21 years of age.   In addition, the questionnaire was modified to automatically reject any participants who did not answer freshman or sophomore as their current year of school.   This second campaign resulted in completed surveys from an additional 33 freshmen and 23 sophomores.   Thus, the total responses from these campaigns consisted of a total of 115 freshmen and sophomores and 101 graduate students.   Counting juniors and seniors, a total of 260 participants completed the questionnaire.   Here again, all junior and senior data were retained for possible later analysis, but were excluded from the analysis of freshmen and sophomore responses.

 

Survey Instrument Design

The survey instrument used in the study consists of six sections.   The first section (administration and demographics) requires entry of a password (online version only) and a code number (affixed to the pencil and paper version).   The participant is then asked to answer five demographic questions (age, gender, year in school, academic major and claimed religious affiliation.   The claimed religious affiliation is included as a check and as an attempt to mitigate the inability for Religious World View Scale to detect religions not falling on a theistic continuum (eg. Buddhist, possibly Muslim, and others).

Section two consists of the Academic Motivation Scale (AMS).   This 5-point Likert scale contains 28 questions dealing with “why I chose to go to college” seems necessarily prior to the remaining instruments.   Section three consists of the Epistemic Belief Inventory (EBI).   The EBI is a 5-point Likert scale with a total of 32 questions.   Section four consists of a 7-point Likert scale with 13 question that constitutes a combined Satisfaction with Life Scale and Subjective Vitality Scale.   Section five consists of the 25 question 6-point Likert scale Religious World View Scale.

Section six consists of two sets of researcher generated questions.   The first set consists of nine Likert scale (7-point) items concerning various aspects of family upbringing (focused on attitudes toward education and religion).   The second set consists of seven Likert scale items (also 7-point) dealing with a few areas of personal beliefs and willingness to independently arrive at these personal beliefs.   This section may allow for some rudimentary, yet direct, comparisons of the domain result to specific beliefs and self reported willingness to engage in worldview integration.

A pilot study consisting of 2 professors, 5 adults (non students) and 5 college students was completed in July of 2005.   Half were requested to complete the questionnaire on paper while half completed it online.   As a result of these comments, the Religious World View Scale was moved from the beginning to near the end of the questionnaire, changes were made to the options available in the last question (Q. 120) and various changes were made to both the paper layout and the online formating.   For example, online questions were paginated to avoid the need for scrolling down the screen.   In addition, the paper version had 5 items for one Likert scale while the online version offered 7 items.   This had to be rectified.   It was also discovered that a 4 point Likert scale in the Religious World View Scale tended to yield a nearly perfect 4.0 mean score among evangelical students.   After changing to a 6 point scale, the average score showed some differentiation.   Additional suggestions involved the wording of various questions within established instruments.   It was decided to avoid making any of these wording changes and to leave the instrument wording intact.   Finally, it was discovered that the majority of participants were able to complete the entire questionnaire in less time than anticipated.   Estimates for the time necessary to complete the questionnaire were reduced from 30 - 45 minutes to 15 - 30 minutes.   The study and questionnaire were approved as submitted by the Protection of Human Rights in Research Committee at Biola University.   No changes to the questionnaire or methodology have been made since that time.

Data Collection

Participants each completed a questionnaire of 120 items entitled “Student Worldview Questionnaire.”   Classroom participants at the evangelical Christian college were given the choice to fill out a pencil and paper version of the questionnaire and to return it to the researcher using a pre-stamped envelope (or intercampus mail) or by completing the questionnaire online over an internet web site.   Participants who completed the questionnaire were offered a few points of extra credit for their class and those utilizing the internet (online) site were provided a coupon (printable) that was good for 20% off of their next purchase from the University Bookstore.   Interestingly, only 12 pencil and paper questionnaires were returned.   The remaining participants all elected to take the questionnaire online.  

The questionnaire was handed out to students in their designated classes as a complete packet (Appendix B) that included the questionnaire, instructions, informed consent form, and an envelope.   The front page of each packet included a password (allowing access to the site) and a numerical code (distinguishing school, class and individual).   The questionnaires were not to be filled out in class, but at the participant’s convenience outside of class.   A notebook was passed around in class for the participants to enter their name, e-mail address and the code number from their packet.   This sheet will be retained for recordkeeping purposes and allowed for the distribution of follow up letters to those who have not yet completed their questionnaire.   This sheet also permitted the instructor to award a few extra credit class points to those who completed the survey.   Informed consent forms were physically signed and returned by those taking the pencil and paper test, and were available for reading and consultation before indicating agreement with the same online.   Three students were exempted from taking the survey because they were under the age of 18.

The participants who were solicited by e-mail campaigns were only able to take the questionnaire online.   The web-based (online) questionnaire was hosted by SurveyZ (www.surveyz.com) and was available 24 hours per day.   The site allowed participants to start their work at one time and then return to it as many times as desired to complete their work.   However, they were not be able to return to previously answered questions and were not permitted to complete the questionnaire more than once or to otherwise “stuff the ballot box.”   Although the questionnaires were identical, after completing the questionnaire, the web page redirected students to different web sites depending upon whether they were students at the evangelical Christian institution or part of the e-mail solicitation.   The Christian institution participants were directed to a printable “20% off” coupon for the institution’s bookstore.   The participants solicited by e-mail were offered the opportunity to provide an e-mail address and receive a return e-mailed coupon for a free movie rental at Blockbuster ®.   The Blockbuster ® coupons were individually generated with help from Blockbuster ® and each contained a bar code that is connected at the store to a real time database.   Copies of coupons with the same bar code will be rejected at the point of sale.

The first classroom participants were asked to complete their questionnaires within a period of 6 weeks.   This proved a bit overly extended and the second group of classroom participants was asked to complete their questionnaires within 2 weeks.   The vast majority of e-mail participants completed the questionnaire within a week of the e-mail solicitation; however responses still drifted in for a period of two more weeks thereafter.   Classroom participants, who had not yet completed the survey, were provide two reminder e-mails sent at two week intervals.   E-mail participants were randomly selected and unknown to the researcher, so follow up messages were not sent.

 

Data Analysis

Data was collected online and downloaded directly from the survey website.   The few pencil and paper responses were then added.   Data was first imported into Microsoft Excel due to the download format and the excellent data organization tools found in Excel.   Descriptive analysis and mean difference tests were performed within Excel.   The organized data was then imported into SPSS – Version 11 (Lead Technologies, Inc.) for correlational and Anova statistical analyses.

The correlational hypotheses, H 0 1.1 to H 0 1.5 will be tested for significant bivariate correlations using the Pearson product-moment test.   A Spearman test will be used if the data is not reasonably within normal distribution patterns and may be used to check for non-linear relationships.   A multiple regression will also used to examine multiple correlations (H 0 1.6).   Hypotheses H 0 2.1 to H 0 4.2 will be tested for significant mean differences using independent t-tests, ANOVA for multiple independent variables, and MANOVA (when appropriate) for multi-dependent variables.   Should significance be found in multiple variable tests and a post-hoc test be required, I will use the Tukey’s test.   This is due to the fact that much of this study is exploratory in nature and the ability to identify possible issues is expected to be more helpful than the need for a more conservative test.

The alpha level will be set at 0.05 for all hypotheses. This level is an appropriate balance that will guard against both Type I and Type II errors by maintaining alpha sufficiently low and power sufficiently high.  Such a level is also not too stringent given both the number of correlations and differences being investigated and the preliminary nature of much of this research.  A two-tailed test will also be employed due to the fact that the direction of the differences and correlations cannot be easily anticipated or supported.

 

Summary

The research portion of this dissertation is largely exploratory in nature and utilizes a quantitative and cross-sectional research design.   It attempts to discover whether any relationships exist between two foundational areas of one’s operational worldview (religious worldview and epistemic worldview or personal epistemology) and the psychological constructs of intrinsic / extrinsic motivation and subjective well-being (vitality and satisfaction with life).   Using a sample of college students from a California evangelical institution of higher learning and a sample of students from a variety of institutions in California, the study was able to collect data from students with a variety of religious and epistemic worldviews.   In addition, each sample includes both gender, a variety of student ages and differentiates between freshman and sophomore students and graduate students.   This will allow the study to also investigate whether the relationships discovered are influenced by main effects of gender, age or year in school.

However, the study faced some unavoidable limitations in obtaining access to students at other institutions and was forced to resort to the use of randomly issued e-mail invitations (with the promise of receiving a free Blockbuster™ rental coupon) to known California students requesting that they take the survey.   Not only did these limitations point out some problems with cooperation among educational research institutions, but this sampling technique resulted in some difficulty in defining the true characteristics of the non-evangelical student population and will yield some limitations regarding generalizing results to that population.

The survey instrument uses a number of well established instruments and was distributed to the evangelical institution in specific classes and could be taken either online or by a pencil and paper questionnaire.   The main difficulty in measurement falls upon the use of the Religious World View Scale, which measures one’s worldview based only upon a continuum of orthodox Christianity.   In addition, the survey instrument contains a number of researcher added questions that should provide some ability to compare the measured concepts to self-reported beliefs and behaviors more directly tied to one’s willingness to engage in critical worldview integration.  

Finally, the use of a predominantly online survey instrument is helpful in the data collection and assimilation process.   Not only does it make the survey more accessible to a variety of students using a format they are generally familiar with, but it also allows data to be downloaded directly into the analysis software.   This helps to prevent error caused by the tedious process of data entry from paper questionnaires.   Once the data is received and organized, the scores on each of these instruments, the individual factors within each instrument, and the results from answers to the final personal upbringing and personal belief questions may be compared and analyzed for significant differences and correlations.   It is hoped that the relationships discovered between these concepts will held us understand how people look at and participate in the worldview formation and integration process.   As an exploratory investigation, this study should provide plenty of data for further thought, consideration and future investigation.

 

Copyright 2005 Mark E. Henze - BBuilders.org