Predictors1
Running Head: Predictors
Temperament, Learning Styles, and Demographic Predictors
of College Student Satisfaction in a Digital Learning Environment
Maribeth Ferguson
CECS 5610
Dr. Gerald Knezek
Abstract
The purpose of this study was to identify predictors of student satisfaction in undergraduate college students at a mid-sized southern university enrolled in courses with a blended learning environment. This study was target sample consisted of all undergraduate students who enrolled in any one of five large-enrollment courses that wereredesigned into blended learning environments. The three surveysadministered to the target sample in this study, the Keirsey Temperament Sorter II (KTSII) (Keirsey, 1998, 2001), the Index of Learning Styles (ILS)(Solomon &Felder, 1993), and the Student Satisfaction Questionnaire (SSQ) Strokes (2003)were requiredassignmentsof the redesigned blended coursessubmitted through an online format. The KTSII, a personality survey, identifies four temperament types: guardian, artisan, idealist, or rational. The ILStransposes results on four dichotomous scales: sensory/intuitive, visual/ verbal, active/reflective, and sequential/global. The SSQdeveloped by Strokes (2003). A reliability coefficient for the scale’s internal reliability with the research samplewas determined using Cronbach’s alpha method. The SSQwas also used to collect demographic data.
The research design used descriptive statisticsto report the demographic data, learning style categories and temperament characteristics of the target sampleat a single point in time. The dependent variable in this research wasthe student satisfaction with the blended learning environment. The Independent variables were temperament, learning styles and demographic characteristics.Data obtained through the SSQ was recorded as interval, ordinal and nominal data. Responses to each satisfaction statement, with the blended learning environment,were reported by using frequencies and percentages for each indicator level. The overall satisfaction score obtained by adding numeric values of the 16 statementsfor each participant ranged from the least satisfaction scoring (16) to the most satisfaction (80). Using the median score as the determinant, the degree of satisfaction wascategorized as unsatisfied or satisfied.
Forward selection of logistical regression was used to predict student satisfaction based on the variables analyzed. The criterion, student satisfaction was correlated with eachprediction variable.The analysis consider two levels of experience with the Internet, novice and intermediate users; and proficient users.The regression equation of the predictor variable on satisfaction indicated whether or not there was a significant effect, and offered the probably of a correct prediction of student satisfaction for the set of predictors. Variables that emerge as predictors of satisfaction were also compared to the individual satisfaction item responses, to identify possible relationships.
The results indicated that gender and internet experience were predictors of student satisfaction in blended learning environments. The significance of this study is in the independent variables that were not predictors of student satisfaction in blended learning environments.
Introduction
Institutions of higher education are turning to online courses as a viable alternative to traditional face-to-face learning. According to Yang and Cornelius (2004),a survey of online education delivered by higher education institutions in the United States, found that at least 80% of the course content delivered by those institutions were delivered online(Allen & Seaman, 2003, as cited by Yang & Cornelius, 2004).The U.S. Department of Education’s survey revealed that more than 54,000 online courses were being offered in 1998 with over 1.6 million students enrolled (Lewis, et al., 1999 as cited by Yang & Cornelius, 2004). The same survey reported that: (a) over 1.6 million students enrolled took at least one online course during the Fall of 2002, (b) of these students 578,000 took all of their courses online, (c) among all U.S. institutions of higher education students in Fall of 2002, 11 percent took at least one online course, and (d) among those students at institutions of higher education where online courses were offered, 13 percent took at least one online course (cited in Allen & Seaman et al., 2003).
At a mid-sized southern university, the goal of their Quality Enhancement Plan, a component of the Southern Association of Colleges and Schools reaffirmation and accreditation process,is to improve student learning outcomes and student experience in large-enrollment undergraduate courses. The University plans to redesign their large-enrollment undergraduate courses into a blended learning format. This five year longitudinal field study will analyzes learner outcomes and learner satisfaction between large-enrollment, traditional, undergraduate courses and blended learning environment courses. As the learning environment shifts from traditional to distance delivery in higher education, the issue of satisfaction with learning emerges (cited by Stokes, 2003). The mid-sized southern university states that 25% of students who enroll in traditional large-enrollment course do not finish the course. It appears that large-enrollment courses are not satisfying students’ needs. The university plans to conduct research to compare learner satisfaction and learner outcomes between the two learning environments.
According to Ascough, (2002as cited by Yang and Cornelius 2004), “Online courses have the following features: (a) they provide a learning experience different than the traditional classroom because learners are different, (b) the communication is via computer and World Wide Web, (c) participation by learners is different, (d) the social dynamic of the learning environment is changed, and (e) discrimination and prejudice is minimized”. According to Paulsen (2002), online education is characterized by: (a) the separation of teachers and learners (which distinguishes it from traditional face-to-face education), (b) the influence of an educational organization (which distinguishes it from private tutoring and self-study, (c) the use of a computer network to present and distribute some educational content, and (d) the provision of two-way communication via a computer network that students may benefit from communication with each other, teachers and staff (p.1).Blended learning is a hybrid of classroom and online learning that includes some of the conveniences of online courses without the complete loss of face-to-face contact.
However, adult learners present a wide range of individual differences including differences in orientation to learning and readiness to learn (Knowles 1970, 1984; Sarasin, 1999 as cited by Stokes, 2003), and no assumptions should be made about adult’s preferencesfor instructional delivery simply because they are adults (Bullen 1997 as cited by Stokes 2003). Keirsey (1998) identified four primary human temperaments: artisan, guardian, idealist, and rational. Felder and Solomon (Felder 1996 as cited by Stokes, 2003) developed the Index of Learning Styles based on Felder-Silverman model in which college students were classified as sensing or intuitive learners, visual or verbal learners, active or reflective learners, and sequential or global learners. Felder promotes the use of multiple instructional approaches to match students’ various learning styles as a means to improve students’ learning, satisfaction with instruction, and self-efficacy. Identifying learner characteristics is basic to studying the use of instructional technology (Gibson, 1998 as cited by Stokes, 2003)
Statement of the Problem
The purpose of this study was to identify predictors of student satisfaction with blended learning environments,according to temperament; learning style; and the demographic characteristics gender, age, grade point average, experience with the World Wide Web, and previous courses taken that incorporated Web-based lessons.Undergraduate students enrolled in the redesigned courses were surveyed to assess their satisfaction with the blended learning environmentto identify possible predictors of satisfaction. The research question was:
Are temperament, learning styles, and demographic characteristics of college students predictors of student satisfaction in a blended learning environment?
Review of the Literature
Distance learning is an increasing important component of higher education (NCES, 1997,1999a, 1999b, as cited in Stokes 2003). Studies have been conducted on the effects of learner satisfaction in an online learning environment. However, few research studies have focused on improving learner satisfaction through a blended learning environment. Recent study involving distance learning has shifted from a focus on the technology itself to its effects on learners. More specifically, this recent research can be classified generally into four categories: interaction, active learning, student perceptions, and learning outcomes (Spiceland 2002).The quality of online educationhas also prompted the attention of higher education accreditation associations (Yang & Cornelius, 2004). In 2000, The Institute of Higher Education Policy (IHEP) first reviewed all of the existing principles or guidelines, and proposed 24 benchmarks for measuring quality Internet-based learning….(Yang & Cornelius, 2004).
New technologies and media have been popular research areas to educators and researchers throughout history of education….While media comparison studies may be useful for making media selection, in fact, they do not contribute much to understanding of technology as a delivery system in education (Clark,1983; Herschback, 1984, McCelland & Saeed, 1986, Moore & Kearsley, 1997; Schrum, 1999, as cited by Yu, Kim &Roh, 2001). Therefore, several factors are crucial for pedagogical decision in the instructional design process (Hajizainuddin 1999 cited by Yu, Kim & Roh 2001) and these factors include learners’ general characteristics, their specific entry behavior, and their learning styles (Heinich, 1996 as cited by Yu, Kim & Roh, 2001).
Petrides (2002 cited by Yang & Cornelius, 2004) conducted a qualitative study to determine learners’ perspectives on web-learning. The research was conducted in a blended university online class, which means the class is a one-semester regularly scheduled class with web-based technology (Learning Spaces) as a supplement. When interviewed, some participants stated, “There is something that forces you to think more deeply about subject areas when you have to respond in writing” (Petrides 2002 cited by Yang & Cornelius, 2004). Students responded that questions in the asynchronous environment allowed more time for reflection than face-to face classroom instruction. Flexibility is an area of strength of the online learning environment that has been identified by researchers (Petrides, 2002;Schrum, 2002 cited by Yang & Cornlenius, 2004). In Petrides (2002 as cited by Yang & Cornelius, 2004) study he report that participants revealed that it was easier to work in collaborative groups in an online course, since there was no needs to rearrange everyone’s schedule (Yang 2004). Convenience was also an advantage reported in the online learning literature. For example, in Poole’s (2002 as cited by Yang & Cornelius, 2004) study of student participation in a discussion-oriented online course, the findings indicated that students participated in online discussions at the times which was most convenient to them, such as on weekends. Poole(2002) also found that students mostly accessed the online course from their home computers, which was the place most convenient to them. Other researchers have also found similar results that online learners read and respond to instructor’s comments in online discussions at times convenient to the e.g. early morning, late evening, (Murphy & Collins, 1997 as cited by Yang & Cornelius, 2004). Proponents of computer-mediated education suggest that the reflectivity, interactivity, and collaboration of online discussion provide an egalitarian learning environment for men and women. Others suggest that on-line discussion contains the same gender bias as face-to-face classroom communication…. It has been suggested that the World Wide Web holds great promise for egalitarian interaction, free from such visual clues as gender, class, social class, and race that may limit of silence speakers. (McAllister & Ting, 2004).
This study used forward selection of logistical regression to predict student satisfaction based on the variables analyzed. Forward selection starts with an empty model. The variable with the smallest P value, when it is the only predictor in the regression equation,was placed in the model. P value—that is, the probability that any particularoutcome would have arisen by chance (Greenhalgh, 1997). Each subsequent step adds the variable that has the smallest P value in the presence of the predictors already in the equation. Small p-values suggest that the null hypothesis is unlikely to be true. The smaller it is, the more convincing is the rejection of the null hypothesis (Eastman, McColl & Young, 1997). Variables were added one-at-a-time as long as their P values were small enough, typically less than 0.05 or 0.10 (Dallal, 2004).
Regression analysis is any statistical method where the mean of one or more random variables is predicted conditioned on other (measured) random variables. Multiple linear regression aims is to find a linear relationship between a response variable and several possible predictor variables (Easton, Hall, & Young 1997). Logistic Regression is a regression method used when the dependent variable is dichotomous. Logistic regression is used to predict the likelihood (the odds ratio) of the outcome based on the predictor variables (called covariates in logistic regression). The significance of the logistic regression can be evaluated by …a Chi-square test, evaluated at the p < .05 level (Lani, 2006).
Assumption in this study include: the students enrolled in the five blended learning courses had the technical skills necessary to participate in a partially Web-based course; the students would understand and answer the surveys honestly; and the target sample would be representative of the institution and the total student population involved in blended learning environments at the postsecondary level.This study’s generalizability of the data is limited, due to the fact; the target sample involved comes from only one institution in the southern United States of undergraduate college students. Additionally, the data is collected at only one point in time. If independent samples are taken repeatedly from the same population, and a confidence interval calculated for each sample, then a certain percentage (confidence level) of the intervals will include the unknown population parameter. Confidence intervals are more informative than the simple results of hypothesis tests (where we decide 'reject H0' or 'don't reject H0') since they provide a range of plausible values for the unknown parameter (Eastman, McColl & Young, 1997).
Method
Research Population and Sample
The target population consisted of undergraduate students at a mid-sized southern university who enrolled in any one of five large-enrollment courses that were redesigned into blended learning environmentbythe institution’s Quality Enhancement Plan, a component of the Southern Association of Colleges and Schools reaffirmation and accreditation process.
Research Design
This study was based on the research design by Stokes, “Temperament, Learning Styles, and Demographic Predictors of College Student Satisfaction in a Digital Learning Environment” (2003). Unlike Stokes, the purpose of this study was to identify predictors of student satisfaction in blended learning environments. The redesigned courses will be taught by the same instructor, delivered by the same department, and require the same content, activities and projects. Like Stokes, (2003), the research design includes a descriptive statistics in which the characteristics of subjects at a single point in time were described. Descriptive items include temperament, learning styles, satisfaction, and demographic characteristics of gender age, grade point average, experience with the World Wide Web, and previous courses taken that incorporated Web-based lessons. The dependent variable in this research was the student satisfaction with learning in a blended learning environment. The independent variables were temperament, learning styles and demographic characteristics.
Data Collection
In this study the three surveys administered to the target sample: the Keirsey Temperament Sorter II (KTSII) (Keirsey, 1998, 2001), the Index of Learning Styles (ILS)(Solomon &Felder, 1993), and the Student Satisfaction Questionnaire (SSQ) developed by Strokes (2003) were required assignments of the redesigned blended courses. The KTSII,a personality survey identifies four temperament types: guardian, artisan, idealist, or rational. The ILS transposes results on four dichotomous scales: sensory/intuitive, visual/ verbal, active/reflective, and sequential/global. The SSQ developed by Strokes (2003), is based on the insights gained from the review of the literature, particularly from Biner, Dean, Mellinger (1999, as cited by Stokes) and Wernet, Olliges, and Delicath (2002 as cited by Stokes). A reliability coefficient for the scale’s internal reliability with the research sample was determined using Cronbach’s alpha method. The SSQ was also used to collect demographic data.
Data Analysis
Like Stokes, (2003) data obtained through the SSQ was recorded as interval, ordinal and nominal data. Descriptive statistics were used to report the demographic data, learning style categories and temperament classification of the target sample. Responses to each satisfaction statement with blended learning environment were reported by using frequencies and percentages for each indicator level. The overall satisfaction score for each participant was obtained by adding numeric values of the 16 statements; the range of possibilities ranges from the least satisfaction scoring 16 to the most satisfaction with a score of 80. The degree of satisfaction was then recoded as unsatisfied to satisfied with the median score as the determinant for the categories.
Also, like Stokes, (2003) forward selection of logistical regression was used to predict student satisfaction based on the variables analyzed. Each predictor variable (temperament, learning style, and demographic characteristics) was correlated with the criterion, student satisfaction variable, with the rating of satisfied or unsatisfied. Two levels of experience were considered in the analysis, novice and intermediate users; and the proficient users. In this study, the regression equation indicated whether or not a significant effect from the variables on satisfaction existed, and offered the probably of a correct prediction of student satisfaction for the set of predictors. Variables that emerge as predictors of satisfaction were also compared to the individual satisfaction item responses to identify possible relationships.
Results
In the Stokes (2003) study, the SSQ was complete by 87% of the participants volunteered to participate in the study, rather than being required to complete the survey. The sample was evenly divided with 52% female and 48% male. Most participants were between 19-23 years old (91%). One half of the students were seniors and one fourth of the students were juniors. The grade classification characteristic would probably be significantly different in this research because the redesigned blended learning courses are primarily lower classman courses. In the Stokes (2003) study, Most of the participants’ grade point averages were between 2.5 and 3.49. Only 2% of the participants defined themselves as novice users. Another 42% defined themselves as intermediate and 56% identified themselves as proficient. One third of the participants indicated they had taken two previous Web-based courses and nearly one fourth reported no experience with Web-based courses. In the sample population for this study those without previous Web-based experience could be significantly larger.