<%BANNER%>

Factors Influencing Program Progression and Degree Completion among Information Technology Students in the Community College

Permanent Link: http://ufdc.ufl.edu/UFE0042475/00001

Material Information

Title: Factors Influencing Program Progression and Degree Completion among Information Technology Students in the Community College
Physical Description: 1 online resource (104 p.)
Language: english
Creator: Jones, Eugene
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: analysis, completion, course, ratio, transcript
Human Development and Organizational Studies in Education -- Dissertations, Academic -- UF
Genre: Higher Education Administration thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The rapid decline of information technology majors poses a serious obstacle for the continued innovation and maintenance of the United States information technology infrastructure. The purpose of this study is to explore barriers to course progression of community college information technology Associate of Science degree students. While the research literature contains several studies about the reasons for the sharp decline among four year undergraduate information technology students very little research has examined community college Associate of Science degree student?s progress toward the information technology degree. A quantitative study using transcript analysis will be conducted to find relationships between course preparation and degree progression among community college students. In addition, logistical regression will be used to determine factors influencing degree completion among information technology students.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Eugene Jones.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Campbell, Dale F.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042475:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042475/00001

Material Information

Title: Factors Influencing Program Progression and Degree Completion among Information Technology Students in the Community College
Physical Description: 1 online resource (104 p.)
Language: english
Creator: Jones, Eugene
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: analysis, completion, course, ratio, transcript
Human Development and Organizational Studies in Education -- Dissertations, Academic -- UF
Genre: Higher Education Administration thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The rapid decline of information technology majors poses a serious obstacle for the continued innovation and maintenance of the United States information technology infrastructure. The purpose of this study is to explore barriers to course progression of community college information technology Associate of Science degree students. While the research literature contains several studies about the reasons for the sharp decline among four year undergraduate information technology students very little research has examined community college Associate of Science degree student?s progress toward the information technology degree. A quantitative study using transcript analysis will be conducted to find relationships between course preparation and degree progression among community college students. In addition, logistical regression will be used to determine factors influencing degree completion among information technology students.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Eugene Jones.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Campbell, Dale F.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042475:00001


This item has the following downloads:


Full Text

PAGE 1

1 FACTORS INFLUENCING PROGRAM PROGRESSION AND DEGREE COMPLETION AMONG INFORMATION TECHNOLOGY STUDENTS IN THE COMMUNITY COLLEGE By EUGENE GARRISON. JONES II A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY O F FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Eugene Garrison Jones II

PAGE 3

3 To Dad, We finally did it!

PAGE 4

4 ACKNOWLEDGMENTS I wish to acknowledge the tremendous supp ort of my committee: Dr. Dale Campbell ( my chair and great supporter), Dr. James Algina Dr. Bernard Oliver and Dr. David Honeyman. Each member pushed sometimes and sometimes patted me on the back to get me through this process. I also want to thank my supporters at Santa Fe College with special thanks to Dr. Anne Kress, Dr. Karen Cole Smith, Dr. Portia Taylor, Dr. Judy Rice, Dr. Dave Yonutas, Kim Kendall and Steve Yongue. I could not even have begun without the help provided by Sa ntosh Kama th in the Of fice Institutional Effectiveness who generated the raw data set. I want to thank Dr. Patricia Grunder for believing in my ability before I was even sure I could do it. To my Mom and the rest of my family I thank you for all the times you keep telling me Most of all I want to thank my wife Keisha for having my back all the way through this happened if not for your love and support.

PAGE 5

5 TABLE O F CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 14 Statement of the Problem ................................ ................................ ....................... 14 Purpose of the Study ................................ ................................ .............................. 16 Significance of the Problem ................................ ................................ .................... 16 Research Questions ................................ ................................ ............................... 18 Methodology ................................ ................................ ................................ ........... 18 Assumptions ................................ ................................ ................................ ........... 19 Limitations ................................ ................................ ................................ ............... 19 Delimitations ................................ ................................ ................................ ........... 19 2 LITERATURE REVIEW ................................ ................................ .......................... 20 Degree Completion ................................ ................................ ................................ 21 Enrollment Trends ................................ ................................ ................................ .. 24 National ................................ ................................ ................................ ............ 24 State of Florida ................................ ................................ ................................ 24 Enrollment at Selected Study Site ................................ ................................ .... 25 Demographic Enrollment ................................ ................................ .................. 25 Information Technology and National Issues ................................ .......................... 25 Financial ................................ ................................ ................................ ........... 25 National Security ................................ ................................ .............................. 26 Factors Influencing Degree Selection ................................ ................................ ..... 27 Self Efficacy ................................ ................................ ................................ ...... 27 Collapse of Dot.com Industry ................................ ................................ ........... 28 Outsourcing of Jobs ................................ ................................ ......................... 28 Media Misrepresentation ................................ ................................ .................. 28 Appeal of Discipline ................................ ................................ .......................... 29 Transcript Analysis ................................ ................................ ................................ 31 Definition ................................ ................................ ................................ .......... 31 Theoretical Framework ................................ ................................ ..................... 31 Summary ................................ ................................ ................................ ................ 32

PAGE 6

6 3 METHODOLOGY ................................ ................................ ................................ ... 38 Research Design ................................ ................................ ................................ .... 39 Population and Setting ................................ ................................ ............................ 39 Setting ................................ ................................ ................................ .............. 39 Participants ................................ ................................ ................................ ....... 40 Access ................................ ................................ ................................ .............. 40 Study Variables ................................ ................................ ................................ ....... 41 Manipulated Variables ................................ ................................ ...................... 41 Measured Variables ................................ ................................ ......................... 41 Dependent variables. ................................ ................................ ................. 41 Independent variables. ................................ ................................ ............... 41 Instrumentation ................................ ................................ ................................ ....... 41 Research Questions ................................ ................................ ............................... 42 Hypotheses ................................ ................................ ................................ ............. 42 Data Collection ................................ ................................ ................................ ....... 43 Data Analysis ................................ ................................ ................................ .......... 45 4 RESULTS ................................ ................................ ................................ ............... 46 Descriptiv e Data ................................ ................................ ................................ ..... 46 Representativeness ................................ ................................ ................................ 48 Use of Models ................................ ................................ ................................ ......... 48 Research Questions ................................ ................................ ............................... 49 Question One ................................ ................................ ................................ ... 49 Hypothesis 1.a ................................ ................................ ........................... 49 Hypothesis 1.b ................................ ................................ ........................... 49 Hypothesis 1.c ................................ ................................ ........................... 50 Question Two ................................ ................................ ................................ ... 50 Hypothesis 2.a ................................ ................................ ........................... 51 Hypothesis 2.b ................................ ................................ ........................... 51 Question Three ................................ ................................ ................................ 52 Hypothesis 3.a ................................ ................................ ........................... 52 Hypothesis 3.b ................................ ................................ ........................... 52 Question Four ................................ ................................ ................................ ... 53 Hypothesis 4.a ................................ ................................ ........................... 53 Hypothesis 4.b ................................ ................................ ........................... 53 Hypothesis 4.c ................................ ................................ ........................... 54 Hypothesis 4.d ................................ ................................ ........................... 54 Hyp othesis 4.e ................................ ................................ ........................... 55 Summary ................................ ................................ ................................ ................ 55 5 DISCUSSION ................................ ................................ ................................ ......... 81 Purpose ................................ ................................ ................................ .................. 81 Comparison with Previous Research ................................ ................................ ...... 82 Comparison 1 Difference in Male and Female Course Completion Ratios ....... 82

PAGE 7

7 Comparison 2 Student Age and the Course Completion Ratio ......................... 83 Comparison 3 Race and the Course Completion Ratio ................................ .... 83 Comparison 4 Grade Point Average and the Course Completion Ratio ........... 83 Discussions of Findings ................................ ................................ .......................... 83 Course Completion Ratio ................................ ................................ ................. 83 Interpretation of the Course Completion Ratio ................................ ................. 84 Question One ................................ ................................ ................................ ... 86 Question Two ................................ ................................ ................................ ... 88 Question Three ................................ ................................ ................................ 90 Question Four ................................ ................................ ................................ ... 92 Limitations ................................ ................................ ................................ ............... 95 Study Sample. ................................ ................................ ................................ .. 95 Geographic Location ................................ ................................ ........................ 95 Implications of the Findings ................................ ................................ .................... 95 Recommendations for Further Research ................................ ................................ 97 LIST OF REFERENCES ................................ ................................ ............................... 99 B IOGRAPHICAL SKETCH ................................ ................................ .......................... 104

PAGE 8

8 LIST OF TABLES Table page 2 1 Intentions of freshmen to major in S&E fields, by race/ethnicity and sex: 2006 .. 34 3 1 Demographic comparisons ................................ ................................ ................. 45 4 1 Group Statistics ................................ ................................ ................................ .. 56 4 3. Descriptive Statistic s ................................ ................................ .......................... 58 4 4. Correlations ................................ ................................ ................................ ........ 58 4 5. Descriptive Statistics ................................ ................................ .......................... 59 4 6 Correl ations ................................ ................................ ................................ ........ 59 4 7 Descriptives ................................ ................................ ................................ ........ 60 4 8 Non STEM ANOVA ................................ ................................ ............................ 61 4 9 Mean Compar isons ................................ ................................ ............................ 61 4 10 Descriptive STEM groups ................................ ................................ ................... 62 4 11 ANOVA STEM group ................................ ................................ .......................... 63 4 12 STEM comparisons ................................ ................................ ............................ 63 4 13 T test Group Statistics ................................ ................................ ........................ 64 4 14 GPA T test ................................ ................................ ................................ .......... 64 4 15 Course Completion Ratio Group Statistics ................................ ......................... 65 4 16 Course Completion Ratio T test ................................ ................................ .......... 65 4 17 Non STEM descriptives ................................ ................................ ...................... 66 4 18 Non STEM Correlations ................................ ................................ ...................... 66 4 19 STEM Descriptives ................................ ................................ ............................. 67 4 20 STEM Corre lations ................................ ................................ ............................. 67 4 21 Descriptive Statistics ................................ ................................ .......................... 68 4 22 Tests of Between Subjects Effects ................................ ................................ ..... 68

PAGE 9

9 4 23 Parameter Estimates. Dependant Variable CCR ................................ ............... 69 4 24 Tests of Between Subjects Effects ................................ ................................ ..... 69 4 25 Parameter Est imates. Dependant Variable CCR ................................ ............... 70 4 26 Dependent Variable Encoding ................................ ................................ ............ 70 4 27 Case Processing Summary Multiple Logistic Regression R esults ...................... 71 4 28 Categorical Variables Codings Multiple Logistic Regression Results ................. 72 4 29 Variables in the Equation Multiple Logis tic Regression Results ......................... 73 4 30 CCR multiplied by 10 Multiple Logistic Regression Results ............................... 74 4 31 All Cases Multiple Logistic Regre ssion Results ................................ .................. 75

PAGE 10

10 LIST OF FIGURES Figure page 2 1 Newly Declared Computer Science Majors ................................ ........................ 35 2 2 Undergraduate Computer Science Enrollment ................................ ................... 36 2 3 Computer Science degrees 1998 2007. ................................ .......................... 37 4 1 Plot of Means, Non STEM gr oups ................................ ................................ ...... 77 4 2 Plot of Means STEM groups ................................ ................................ ............... 78 4 3 Box Plot of GPA ................................ ................................ ................................ .. 79

PAGE 11

11 LIST OF ABBREVIATION S Algorith m A logical sequence of steps for solving a problem, often written out as a flow chart that can be translated into a computer program AA The Associate of Arts ( AA) is the community college degree in the state of Florida designed for university transfer. I t is covered by the statewide articulation agreement, and students completing an AA degree at a community college are guaranteed transfer to a state univer sity. AS The Associate of Science (AS) is the community college degree in the state of Florida design ed for workforce employment. The curriculum frameworks are aligned to the Career Clusters delineated by the U.S. Department of Education Community College At the time of this study the state of Florida has granted permission for community colleges to offe r the baccalaureate degree. Therefore many community colleges in the state of Florida are changing their name dropping the term community from their name Course completion Course completion is li mited to successful completion of a course for an actual grade, it is also not considered a successful completion. Course com pletion ratio The ratio is comprised of the number of hours successfully completed (numerator) and the number of hours attempted (denominator). District Geographic boundary comprised of one or more counties. Florida community colleges serve discrete di stricts comprised of counties identified by the state. High schools within the service district are First time in college (FTIC) Student population limited to those students who have not previously enrolled in postsecondary education. A FTIC student may bring with him or her credits earned by exam (e.g., AP credits) that may be taken in high school prior to entry to college. Full time Enrollment status is at least 12 credit hours in a major (fall or spring) academic semeste r. Enrollment for at least 20 credit hours in an academic year (12 fall, 12 spring, 6 summer) is considered full time on an annual basis for this study.

PAGE 12

12 Gatekeeper A course that must be successfully passed in other to continue in the degree program sequ ence. Usually this course has either a high repeat rate or a high failure rate on first attempt. GPA The Grade point average GPA is calculated on a cumulative basis. In each calculation, the following numerical assignments hold true: A = 4, B+ = 3.5, B = 3, C+ = 2.5, C = 2, D+ = 1.5, D = 1, F = 0, WF Non continuous enrollment Student enrollment pattern interrupted by a period of one semester or more in which a student does not register for courses. Part time Enrollment status is fewer than 12 credit hours in a major (fall or spring) academic term; enrollment in fewer than 20 hours annually is considered part time on an annual basis. Remedial course A course is offered at the community college for subject matter remediation that does not carry college level credit nor count toward an earned degree or certificate. Students are placed into specific levels of developmental work based upon their College Placement Test (CPT) scores. In Florida, minimum placement scores for remedial course s are assigned at the state level; remediation is available in three areas: mathematics, reading, and writing. STEM A program major in the Science, Technology, Engineering or Math studies group. Withdrawal A course not completed by a student. A student e (withdrawal failure) if she or he withdraws from the course after the last day to withdraw or earn a no the student transcript data set

PAGE 13

13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FACTORS INFLUENCING PROGRAM PROGRESSION AND DEGREE COMPLETION AMONG INFORMATION T ECHNOLOGY STUDENTS IN THE COMMUNITY COLLEGE By Eugene Garrison Jones II December 2010 Chair: Dale F. Campbell Major: Higher Education Administration The rapid decline of information technology majors poses a serious obstacle for the continued innovati on and maintenance of the United States information technology infrastructure. The purpose of this study is to explore barriers to course progression of community college information technology Associate of Science degree students. While the research lit erature contains several studies about the reasons for the sharp decline among four year undergraduate information technology students very little research has examined community college Associate of Science degree progress toward the information technology degree. A quantitative study using transcript analysis will be conducted to find relationships between course preparation and degree progression among community college students. In addition, logistical regression will be used to determine fa ctors influencing degree completion among information technology students

PAGE 14

14 CHAPTER 1 INTRODUCTION According to data published by the Computer Society Connection the percentage of undergraduates indicating they would be seeking a degree in computer scien ce or related field dropped by 70% from 2000 to 2005 (Ward, 2007). This is a perplexing decline considering that information technology jobs continue to rise. In the state of Florida information technology job growth is so important that it is listed as one of the Targeted Industry Sectors by Workforce Florida Inc. According the U.S. Bureau of Labor Statistics, computer analysis and computer programming will continue to be in the top 50 of jobs with the most openings nationwide from 2006 to 2016. The wi dening gap between the number of job openings and qualified graduates to fill them not only has economic impact but national security implications as well (Gibson, 2005). Another alarming effect of the reduction in students pursing information technology is the skewed impact on minorities entering higher education. The current decline in enrollment has been forecast by Tucker (1999) and Ward (2007). Also, Ntiri (2001 ) has pointed out that the lack of knowledge of technology causes a barrier to successful ly pursing higher education. This idea combined with belief that the lack of cultural diversity in instructional design results in the production of software that is culturally biased positions minority students at steep disadvantage. Statemen t of the Problem The field of information technology has emerged to become the catalysis for the global economy (Akinson & McKay, 2007). Every aspect of modern society is welded to information technology in the design, production, transportation, marketin g or sale of goods and services. According to data from the U.S. Department of Labor, the demand

PAGE 15

15 for information technology skilled workers will continue to rise and outpace the supply of workers in the field. Because information technology being used fo r the current to the very structure of modern society. Gilman (2009) reports that revenue created by illegal use of information technology is in the range of 10 bill ion annually and the global damage caused by criminal cyber activity is estimated by the U.S. Federal Bureau of Investigation to be around 100 billion. The alarm is also being raised concerning the threat to U.S. national security. Popp(2009) and Clark & Wesly(2009) state that the new battle field is in the computer servers, routers and hard drives of the international competitors. Development of a highly skilled workforce must become a priority for the United position as a global leader. Juxtaposed against the evident need is widening gap of students entering the information technology programs. Gibson (2005) reports that nationally enrollment in information technology programs are down 60%. The alarm of thi s plight has been raised by researchers in the past years. Adelman (1997) stated almost 14 years ago that the educations system must be empowered to recruit, develop and graduate larger numbers of students in the information technology fields. To increas e graduation numbers two approaches can be taken. The first is to increase the number of students in the program via increased recruitment efforts. The second method is to increase retention of the students currently enrolled in the program. Tinto (1993 ) found that student involvement leads to increased retention and enhanced progression toward graduation. There also exists research that points to the relationship between student demographics and course

PAGE 16

16 enrollment patterns as indicators of progress towa rd graduation. Adelman (2004) presented research that the course completion ratio is a valid measure of degree progression. Purpose of the Study The purpose of this quantitative transcript analysis study is to identify demographic course enrollment factor s that influenced course completion percentage of the information technology degree among a select group of community college students. This study will also identify factors that influence degree completion for information technology students in the commu nity college. While there are many studies that describe the issue of four year undergraduate enrollment decline in technology fields, ( Denning & McGettrick, 2005 ; McBride, 2007 ; Foster, 2005 ), a review of the literature reveals only a small number studie s that have directly studying community college students. Two studies directly researching the role of the community college in technology education were Lerman, Riegg, and Salzman, ( 2001) and Lerman, Riegg, and Salzman, ( 2000). However at the time of pu blication the computer technology field was not experiencing the rapid decline it is currently facing, therefore those articles do not address the reasons for declining enrollment. Significance of the Problem The role information technology plays in the ec onomic, financial, telecommunications and national security of the county can not be understated. Atkinson & McKay (2007) have stated that information technology is the driver of not A study commissioned by the Joint Economic Committee (2001) asserted that continued

PAGE 17

17 continued national economic growth. The rapid development of information techno logy also has fundamentally changed the approach and development of military tactics. Berkowitz (1995) stated that the United States is more dependent on electronic information technology systems than any other developed country. Because of this, it is ea sier and cheaper to launch potentially devastating information technology attacks to civilian livelihood than other methods. Compared to assassinations, hijacking, sabotage and explosives information technology is less risky and potentially far more effec tive in disrupting key infrastructure systems such as banking, electrical grids and telecommunications (Berkowitz, 1995). The financial infrastructure also is dependent on information technology to ensure that financial services will continue to operate a t desired levels. The Office of on computer software trading algorithms could potential cause artificial swings in the market trading. (Office of Technology Assessment, 1990). This dissertation seeks to employ transcript analysis and logistical regression information technology major. The study explores potential barriers to student p rogression of the information technology degree and seeks to provide demographic profiles of successful students who enter and finish the information technology degree. The data presented in this study will give insight into student degree progress and al low recruitment and retention plans to be based on empirical data of student progress and performance.

PAGE 18

18 Research Questions 1. Is there a significant relationship between selected student demographics (i.e. race, gender, age) and course completion ratios betwee n program majors? 2. Is there a significant relationship between college grade point average, program major and course completion ratio 3. Is there a significant relationship between student enrollment patterns and course completion ratio 4. Is there a signifi cant relationship between elected student demographics (i.e. race, gender, age, STEM, course completion ratio ) and likeli hood to Graduate? Methodology This study was conducted on transcript and student record data from the internal data warehouse of the s tudy site college. The data was comprised of records from 1,292 first time in college students. The study grouped students into two groups; those seeking science, technology, engineering and math (STEM) degrees and those who were not. The study compared the college transcripts to investigate factors influencing progression through the program and program graduation. The study followed the students from Fall 2003 to Spring 2010. The study used demographic and academic data to compare and contrast studen ts who chose STEM programs versus those students who did not. Comparisons were made across several demographic factors such as race, gender and age. The study also compared the two groups in performance areas such as grade point average and course comple tion ratios. Finally the study used multiple logistic regression to investigate what factors were predictors for program graduation for the two groups. The purpose of the study will be to determine factors that influence program progression and degree com pletion of Information Technology degree among

PAGE 19

19 community and technical college students. While the research focus is on Information Technology students the study used a wider data set of program majors. This approached was done because many students chan ge degree programs with STEM and the viable set of data demographics would apply across STEM program groups. Assumptions For this study, the following assumptions were made: 1. The measures were reliable and valid indicators of the constructs measured. 2. The purposes, processes, and elements of the framework studied have a degree of applicability and generalize to community colleges with similar student populations and similar close proximity to state universities. 3. The primary data were accurately recorded. L imitations This study was limited to data collected on cohort data collected from Fall 2003 to Spring 2010. This study was limited in student data was collected from one community college in Florida. The validity of the study was limited to the reliability of the instruments used. Delimitations This study confined itself on the cohort data of students who entered the study college in Fall 2003 and the student records until the Spring of 2010. This study used student demographic data (race, gender, age) as student group identifiers. This study used enrollment data (course name, program major, grade point average, course completion ratio) for analyzing course progression toward degree. All first time in college students who enter the study college in Fall 2003 were included in this study.

PAGE 20

20 CHAPTER 2 LITERATURE REVIEW This chapter is divided in 2 sections to provide an overview of the research literature that supports the need for this study. Section one addresses the current interest and national focus on degree completion among community college students. The section 2 has five subparts that explore reasons as to why degree interest and completion is so low among Information Technology students. Section 2 part one explores enrollment trends that detail r apid decline of information technology students. Section 2 part two addresses the demographics of current information technology students. Section 2 part three offers a review of the literature detailing the importance of information technology innovatio n and growth to several national issues. Section 2 part four explores the internal and external factors that are influence student entrance to information technology studies. Section 2 part with the framework of t he course completion ratio and transcript analysis as an effective and reliable means of prototyping student characteristics and demographics in terms of course selection and degree completion. Due to the wide influence of information technology in not on ly academia but society as a whole, literature sources are used from a variety of disciplines and fields. Examination of the research literature reveals several interchangeable terms that all refer to the field of computer science or information technolog y. Unlike other fields such as biology or mathematics that have existed for centuries, computer science is still evolving and therefore the nomenclature has not been fully cemented. Denning & McGettrick (2005), identify the following academic programs as related to computer science; computer engineering, information technology, information sys tems and software engineering.

PAGE 21

21 Degree Completion Currently several national initiatives exists that focus on degree completion among community college students. In 2010 the Obama administration stated to two national goals: by 2020, increase the number of college graduates to a level that establishes the United States of America as the country having the highest number of college graduates in the world, and to increa se the number of community college graduates by 5 million. In 2010 the Whitehouse also held the first ever summit on Community Colleges to highlight the important role Community Colleges play in the education of American students and the need for increase d resources to be used to increase the number of community college graduates. The Bill and Melinda Gates Foundation has also launched a major initiative to increase college degree completion. The Foundation has produced a concept paper based on the rese arch of Goldbrick Rab(2007) which uses the Loss and Momentum framework. This conceptual frame work states there are four key points connection, entry, progress, and completion valid measurements must be taken and evaluated at each point. The theoretical concept of measuring academic momentum is also validated by Adelman(2005) and is used in this study as a theoretical framework. The College Board Advocacy and Policy center released in 2010 The College C ompletion Agenda, a national plan to increase the number of 25 to 34 year olds who hold an associate degree or higher to 55 percent by the year 2025 in order to make America the leader in educational attainment in the world. Currently the United States r anks 12 th out of 36 nation in educational attainment for 25 to 34 year olds. Critical to this goal is to increase the number of low income and minority student graduates. In

PAGE 22

22 order to reach the 55% goal stated by the commission, over half of the new gradu ates must come from current low income and minority students. The results of this study include findings on how well minority students progressed in the study cohort. This study analyzes degree completion in the framework of academic momentum. Ad elman(20 05) defines academic momentum as the rate a student is progressing toward their degree. A student can not complete their degree unless they complete all necessary course work. But Adelman found that the rate at which a student progresses influences the o dds of whether a student ever eventually completes. Therefore, identifying measurements of student academic progression, momentum, is critical to understanding the likely hood of degree completion. Goldbrick Rab(2007) also validates the measurement of aca demic momentum as frame work to study degree completion. Goldbrick Rab establishes four key points that students should be tracked and charted to understand student progression toward completion. In the Loss and momentum framework, Goldbrick Rab states t hese four points are; connection, entry, progress, and completion The points mentioned are actually stages that have time frames they may last years. During each stage Goldbrick Rab suggests measurements be used to gauge to amount of momentum the studen t is gaining to ward moving to the next progression point Perrakis(2008) studied degree completion factors among African American and White male community college students. Perrakis used transcript analysis to follow a cohort of 4,333 students and conduct ed regression analyses to find factors influencing degree completion. The findings revealed that highest level of math completed was the

PAGE 23

23 largest influence on degree completion. This study also validated the use of transcript analysis as a tool to study d egree completion.

PAGE 24

24 Enrollment Trends National Several studies have documented the rapid decline in students majoring in information technology. Ward (2007) summarized a report from the Computer Society Connection stating that enrollment in information tec hnology courses dropped 70% from 2000 to 2005 Babco, Zumeta & Raveling (2000) warned that the national trend of students not pursing science and engineering will severely impact the nations ability to continue to keep its leadership in the global economy. Vegso (2008) reported that the declining trend continued from 2003 2004 to 2006 2007 with enrollment in undergraduate information technology courses falling 18%. A graphical summary of the enrollment decline presented by Vegso appear s in Figure 2 1 ,Figu re 2 2 and Figure 2 3. While the data reported certainly highlight the rapid national decline among undergraduate students, the report also supports the need for this study. The current literature is shallow in tracking the selection and persistence of c ommunity college information technology students. State of Florida When the enrollment of information technology students is examined in the state of Florida the percentage of majors is even higher than national percentages. The Florida Community College Fact Book reports that in 2004 2005 only 1,446 FTE students enrolled in Computer Science & Information Technology system wide. Total FTE enrollment system wide is 294,818. Therefore, information technology students made up only 0.4 % of system wide enrol lment. The 2007 edition of the Florida Community College Fact Book reports only 1,409 FTE students enrolled in Computer Science & Information Technology system wide. Total FTE enrollment system wide is

PAGE 25

25 244,720. Therefore, information technology students made up only 0.5 % of system wide enrollment. Clearly the progression path of information technology students from the community college to the university is nearly non existent. Enr ollment at Selected Study S ite The student data warehouse of the selecte d study site reports the enrollment figures for information technology students for Fall 2008 to be 283 while the total college enrollment was reported as 15,649. The ratio of information technology students to the total enrollment is 1.8%. Demographic En rollment While low numbers of students pursing information technology careers has been documented in this study, when factors such as race and gender are considered the low en rollment is dramatic. Table 2 1 illustrates the enrollment break down for all un dergraduate programs in 2006 in Science and Engineering programs based on a survey on freshman intended majors. Across all minorities only 1.6% of students indicate an intention to major in information technology. This has serious implications in the fut ure design of software and instructional software in particular. Perkins (2008) has stated that there is a need for instructional designers from various backgrounds to ensure the provision of culturally sensitive instructional software. Chin (2007) stat ed that the lack diversity of software programmers for instructional design students will be also become a disadvantage in the learning environment. Information T echnology and National Issues Financial In remarks made at the Journal of Financial Services R esearch and the American Enterprise Institute in 2000, Allan Greenspan stated the vital importance of information

PAGE 26

26 not see any end to the continuing innovation and finan cial development that information Atkinson & McKay (2007) claimed that information technology is the driver of not commiss ioned by the Joint Economic Committee (2001) reported that continued continued national economic growth. The financial infrastructure also is dependent on informatio n technology to ensure that financial services will continue to operate at desired levels. The Office of Technology Assessment issued a report in 1990 that potential cause ar tificial swings in the market trading. (Office of Technology Assessment, 1990). National Security The rapid development of information technology also has fundamentally changed the approach and development of military tactics. Berkowitz (1995) states that because the United States is more dependent on electronic information technology systems than any other developed country that it is easier and cheaper to launch potentially devastating information technology attacks to civilian livelihood than other meth ods. Compared to assassinations, hijacking, sabotage and explosives information technology is less risky and potentially far more effective in disrupting key infrastructure systems such as banking, electrical grids and telecommunications (Berkowitz, 1995)

PAGE 27

27 Factors Influencing Degree Selection Currently the literature presents five themes that shape the degree selection of information technology students. These themes are: S elf efficacy ( do they feel they can successfully complete the major of their choice ) the collapse of the dot.com industry ( do the students feel the industry as a whole has stopped expanding ), outsourcing of jobs ( do the students feel most jobs are sent over seas), misrepresentation of job market by mass media (how has the media convince d current students not to pursue information technology careers) and the appeal of the discipline (the perception that the field is boring and unappealing). Self Efficacy d efficacy can play a major role in how they approach goals, tasks, and challenges in academic settings (Bandura, 1993). Both how well they thought they would perform in computer related classes greatly affected whether they attempted to enroll in those classes. The use of self efficacy must be done carefully because of the complex nature of this construct. Kolhede (2001) found successfully complete certain degree programs Fountain (2004) found that females self efficacy beliefs in perform ing well in computer science were greatly influenc ed by how welcoming they perceived the department academic program to be. In relation to minority students self efficacy plays a more complex role in information technology pursuits. Allen(1991) noted that black students striving to purse academic achiev ement at predominantly white colleges have lower levels of self efficacy than white students.

PAGE 28

28 When this is combined with increase stress of the challenging nature of information technology curriculum, black students are at higher risk to not completing th e degree. Collapse of Dot.com Industry The popular notion that the burst of the dot.com era has influenced thousands of students is echoed in many research articles, ( Anthes, 2006), ( Arora & Chazelle, 2005; Holmes, 2007) and ( Granger, Dick, Jacobson, & Van Slyke, 2007 ). Hayes (2005) reported that the two events, the collapse of dot.com and the outsourcing of jobs, are interrelated and influence each other. The creation of the surge of outsourcing was a direct result of the collapse of the many internet st artup companies (Hayes, 2005). Outsourcing of Jobs Brotheron (2003) found that the belief that most information technology jobs are being out sourced was a major deterrent to students seeking information technology careers. The reality is that most high l evel information jobs are not being outsourced and of those jobs that are 50% end up being relocated back to the U.S. (Hayes, 2004). Holahan (2007) stated that if the United States was producing enough information technology workers there would be no need for the outsourcing trend even with the lower wages of over seas workers. Media Misrepresentation Anthes (2006) conducted an interview of 6 faculty members employed in questione d them on the current state of declining student interest in computer science. Several faculty stated that the public media is scarring students off with tales that there are no jobs to be had. The challenge faced by the advocates of information technolog y careers, such as the faculty mentioned, is that the popular media does not report the

PAGE 29

29 readily available data that show definitively that job openings will continue to increase in technology fields. Murphy (2005 ) urged corporations to partner with academ ic institutions to present a positive image to the media about the careers in information technology. Tietjen ( 2004 engineering fields (including information technology) is the huge negative imag e problem that the field has. Until the general public can be educated perhaps through the media about what the career field really has to offer and what members of the profession actually declining enrollment is likely. Appeal of Discipline Of the five c authors stated that if information technology is to survive it is up to the faculty to save it (Arora & Chazelle, 2005). Information technology faculties have little influence over the other constructs mentioned in the literature. Self efficacy is by definition an internal mental element, faculty can not control the media outlets nor can they dire ctly control business decisions such as out sourcing or industry collapses. But they do have a wide range of control over the curriculum and how the information technology programs are marketed to students. Akbulut & Looney (2007) proposed a vocational ps ychology model that suggests that interest, self efficacy, and outcome expectations influence a student's choice of major. By following this model the authors state interest in information technology can be created among the student body. The authors eve n suggest self efficacy can be developed through opportunities for students to achieve and curriculums that incorporate current content can increase interest in computer courses. Arora & Chazelle (2005) stated a collective failure among educators,

PAGE 30

30 practit ioners and researchers to present a compelling narrative about computer science that would excite students. The authors further stated that the high school advanced placement tests for computer science are all about programming and the focus in high schoo l is on programming languages rather than higher level concepts. Another marketing misstep is allowing for mathematics to dominate the computer science examples. Students often associate computer science with mathematic studies ( Campbell, 1992). Denning & McGettrick (2005) stated that the computer science profession has allowed itself to be portrayed in a very narrow view of just programming Forte & Guzdial (2005 ) echoed this by stating introductory courses often fail to connect programming with studen future courses. Also unpopular among students is the idea that computer science equates to being located in a cubicle hunched over a terminal for 12 hours a day ( Forte & Guzdial, 2005). The members of the computer science profession are very frank in their assessment that the way computer science courses are taught is not engaging to curriculums are putting students to sleep by asking them to program simple, tedious tasks ( Anthes, 2006). Arora & Chazelle (2005) noted that students should be taught to view computer science in terms of the great accomplishments the field has made on society such as the coding of the human genome or the algorithms that make Google search so powerful. Holmes (2007) feels teaching computer science as primarily programming is a huge mistake and dilutes the true calling of the field. Ayala & Striplen (2002 ) state that a more focu sed effort should be done to create career introduction programs for first generation students. A program of such nature would help give an

PAGE 31

31 opportunity to discuss the many opportunities that exist in information technology to a group of students that may not have been exposed to the career Kohede(2001 ) found gender differences factor into degree selection but offers how the marketing of the degree could impact degree selection. Soh, Samal & Nugent(2007) stated the fundamental approach to teaching introd uctory information technology courses is flawed and presented 10 general design strategies to redevelop curriculum. Transcript Analysis Definition Transcript Analysis can be define d as the coding and use of enrollment files, college application data, finan cial aid records, and other data that community colleges must routinely collect to comply with state and federal reporting mandates (Hagedorn & Kress, 2008) Theoretical Framework This study is uses the framework of Academic Momentum as defined by Adelman(1 999,2005). Adelman proposes that student progression can be measured by the amount of momentum the student generates in their course work. The concept of momentum can be conceptualized in the Course Completion Ratio(CCR) measurement. The CCR is defined the total number of hours attempted. Calculating the CCR each semester will give the amount of momentum the student is making academically toward their degree. The CCR has been used in seve ral studies Hagedorn(2004), Kress(2007) and Hagedorn and Kress(2008). These studies have used transcript analysis as the methodology to generate the Course Completion Ratio. The foundational work that establishes transcript analysis as effective methodol ogy was produced by Adelman in a series of

PAGE 32

32 longitudinal transcript studies (1990, 1992, 1994, 1997, 1998, 2000, 2004). The use of transcripts as an objective data source has influenced the state of Florida to adopt dation of its own state wide community college data warehouse (Wyman, 1997). The Transfer and Retention of Urban Community College Students (TRUCCS) study, based in the Los Angeles Community College District, demonstrated the value of the transcript analy sis framework in the study of the retention, completion and transfer of community college students (Hagedorn, 2005; Maxwell & Moon, 2001). The validation and reliability of transcript analysis has been verified and confirmed in studies conducted by Garris on, Clevland Innes, Kooke & Kapleman (2006) and Hagedorn, Adelman & Chen (1998) Summary It is clear much has been written to address the problem of the decline in computer science enrollment among four year undergraduate students. But as noted earlier, t here is very little published concerning if these same reasons apply to the community college student. Data from the U.S. Department of Education profile the community college student as being older, usually self supporting and more racially diverse than the profile of the typical four year undergraduate. According to the 2006 Florida Department of Education Fact Book most community college A.S. program students, including information technology, already have a degree it in another field. As such it would seem that these students would be more likely to have researched job prospects and would be aware of the expansive job openings. However, data from the 2000 and 2006 Florida Department of Education Fact Books reveals that there is a steady decline in com munity college students seeking the information technology degree. The literature has not revealed why community college A.S. students are not

PAGE 33

33 pursuing the information technology degree. This research study will give a voice to the often overlooked popul ation of the community college A.S. student and what process they use in degree selection. This study will also add to the research literature and give insight to program directors and advisors of A.S. students allowing those advisors to better guide toda

PAGE 34

34 Table 2 1 Intentions of freshmen to major in S&E fields, by race/ethnicity and sex: 2006 (Percent) Race/ethnicity and sex All S&E majors Biological/ agricultural sciences Computer sciences Engineering Math ematics/ statistics Physical sciences Social/ behavioral sciences All races/ethnicities 32.0 8.8 1.6 8.0 0.8 2.3 10.5 Female 27.2 9.6 0.4 2.5 0.6 2.0 12.1 Male 37.9 8.1 3.0 14.5 0.9 2.8 8.6 White 29.5 7.5 1.5 7.6 0.8 2.5 9.6 Female 24.2 8.1 0 .3 2.2 0.7 1.9 11.0 Male 36.4 7.0 3.0 14.2 0.9 3.0 8.3 Asian 44.7 17.3 1.5 11.7 1.1 2.8 10.3 Female 39.4 18.7 0.5 4.6 0.9 2.5 12.2 Male 50.7 15.3 2.8 19.7 1.3 3.3 8.3 Black 34.0 9.4 2.2 7.2 0.5 1.6 13.1 Female 32.1 11.0 1.1 2.3 0.4 1.6 15.7 Male 36. 3 6.8 3.8 14.4 0.8 1.8 8.7 Hispanic of Mexican/Chicano/Puerto Rican descent 35.9 9.8 1.2 6.4 0.9 2.1 15.5 Female 32.5 10.1 0.1 2.1 0.7 1.6 17.9 Male 40.8 9.2 3.1 13.5 1.1 2.6 11.3 Other Hispanic 36.7 10.0 1.3 7.6 0.6 1.8 15.4 Female 34.2 10.6 0.3 2.8 0.6 1.5 18.4 Male 40.1 8.9 2.8 14.6 0.9 1.9 11.0 American Indian 34.4 10.0 1.6 7.5 0.4 2.7 12.2 Female 30.8 11.9 0.4 2.4 0.2 2.4 13.5 Male 39.5 7.7 3.3 14.2 0.7 3.2 10.4 NOTE: Includes first year students at all 4 year colleges. Race/ethnicity catego ries are those used in the survey's data collection.

PAGE 35

35 Figure 2 1. Newly Declared Computer Science Majors

PAGE 36

36 Figure 2 2. Undergraduate Computer Science Enrollment

PAGE 37

37 Figure 2 3. Computer Science degrees 1998 2007.

PAGE 38

38 CHAPTER 3 METHODOLOGY This study e mployed a quantitative approach to research by using transcript and student record data from the internal data warehouse of the study site college. The study analyzed degree completion as framed in the concept of academic momentum. Academic momentum was defined in the course completion ratio measurement. The methodology approach was a longitudinal study of a student cohort using transcript analysis. The data was comprised of records from 1,292 first time in college students. The study grouped students into two groups; those seeking science, technology, engineering and math (STEM) degrees and those who were not. The study compared the college transcripts to investigate factors influencing progression through the program and program graduation. The stud y followed the students from Fall 2003 to Spring 2010. The study used demographic and academic data to compare and contrast students who chose STEM programs versus those students who did not. Comparisons were made across several demographic factors such as race, gender and age. The study also compared the two groups in performance areas such as grade point average and course completion ratios. Finally the study used multiple logistic regression to investigate what factors were predictors for program gr aduation for the two groups. The purpose of the study will be to determine factors that influence program progression and degree completion of Information Technology degree among community and technical college students. While the research focus is on Inf ormation Technology students the study used a wider data set of program majors. This approached was done because many students change degree programs within STEM and the viable set of data demographics would apply across STEM program groups.

PAGE 39

39 Research Desi gn The design for this study was a longitudinal transcript analysis. There exists an influential body of research including Adelman (2004) and Hagedorn (2005) that establishes transcript analysis as a research methodology and as such it serves as the meth odology approach for this study. This study used two statistical phases. In the first phase comparative statistical analyses were used to explore if differences existed between the STEM and Non STEM students. In this phase independent samples t tests an d analysis of variance techniques were employed. In the second phase logistic regression was used to explore influences on the likely hood to graduate between STEM and Non STEM students. The subject of this study emerged from a review of the literature o n the emerging concern for the lack qualified information technology works and the decreasing flow of students graduation in the information technology field. This study is based in research concerning factors that influence degree completion and uses the course completion ratio as a valid construct in progression to degree. Logistical regression models are used to predict STEM degree graduation. Descriptive and inferential statistics are used to examine and analyze relationships between variables identi fied in the study. Regression modeling is used to compute correlations on all applicable variables. Population and Setting Setting The retrospective data for this study was accessed from a community college located in the North Central area of Florida. T his college is a comprehensive, public two year institution that serves a two county service district. The college serves approximately 16,000 students the majority of which are university transfer.

PAGE 40

40 Approximately 3,700 students are enrolled in technical and vocational programs. The following data describes the college in the Fall of 2009. Total student headcount was 17,391 comprised of 9,115 females (52.4%) and 8,273 males (47.6%). The number of students who registered for a part time load, less than 1 2 hours, was 9,345 and the number of students register for 12 or more hours was 8,046. The two largest student minorities are African American with 2,930 students (16.8% of population) and Hispanic with 1.744 (10.0% of population). The two largest studen t age groups were 15 to 19 with 6,071 students(34.9% of population) and 20 to 24 with 5,847 students(33.6% of population). Of the students who come to the college, 52.8% come from the two county me from other counties in Florida. The college has s even educational centers but only one comprehensive center that offers al l the resources such as library, gymnasium and food court. Table 3 1 shows a comparison between the study college, the state of F lorida and national demographics. Participants Data was collected from Fall 2003 Spring 2010 for all first time in college students admitted to the college. All personable i dentifiable data were de identified by the Access research office. Access to student record data was obtained from the Provost of the college.

PAGE 41

4 1 Study Variables Manipulated Variables This study did not include any manipulated variables. Measured Variables Dependent variables. the total cred it hours in which a student initially enrolled Graduated: Contains 1 if student graduated by Spring 2010 or 0 if student did not graduate Independent variables. STEM group: Indicates whether student chose a science, technology, math or engineering program as a degree or not. Contains either 0 or 1 Age enrolled : 3 Average Term Enrollment: average semester hours enrolled over all terms of active enrollment in the study time frame Race: Caucasian, African American, Hispanic, Asian/Pacific Islander, American Indian/Alaskan Native and Other (for a category not defined by any other available) Gender : male/female Gatekeeper course : Either ENC1101, College Composition or MAT1033, Intermediate Algebra. Grade Point Average: A unit measure of course grade points divided by course hours. Instrumentation The research base for using transcript analysis is from the work of Clifford Adelman who is currently a researcher for the Institute for Higher Education P olicy and who served as research analyst for the Department of Education for over 30 years. Adelman (2006) made the claim in his recent work on the path from high school to

PAGE 42

42 college that studying the high school pattern is essential in predicting the succe ss of college completion. McCormick & Carrol (1999) used transcript analysis in a study of beginning students at four institutions. This study will use the methods of Group Parsing and Group Comparisons that were employed by Hagedorn (2005). The use of re gression analysis to predict degree completion has been analyzed in studies by Koefoed (1984) and Lightfield & Rice (1974). Research Questions 1. Is there a significant relationship between selected student demographics (i.e. race, gender, age) and course com pletion ratios between program majors? 2. Is there a significant relation ship between college grade point average program major and CCR. 3. Is there a significant relationship between student enrollment patterns and CCR. 4. Is there a significant relationship betw een elected student demographics (i.e. rac e, gender, age, STEM, CCR) and likeli hood to Graduate? Hypotheses 1.a: Male students will have higher CCR percentages than female students in STEM programs. 1.b: Students who begin enrollment at older ages will hav e higher CCR than younger peers in STEM programs. 1.c: Black students will have lower percentages in CCR in STEM programs than all other student peer groups. 2.a: Students in STEM majors will have a higher gpa and course completion ratio when compared to Non STEM students 2.b: STEM students with higher grade point averages will demonstrate higher course completion ratios 3.a: Higher credit hour enrollment positively impacts course completion ratios 3.b: Enrollment in either a gatekeeper course or an academ ic foundation course negatively impacts course completion ratio

PAGE 43

43 4.a: Female students enrolled in STEM programs have higher odds of graduating than male peers 4.b: Older students are more likely to graduate than younger students in STEM programs 4.c: Cours e completion ratio is a significant predictor of whether a student graduates or not in STEM programs. 4.d: Being enrolled in a STEM program is a significant predictor of graduating 4.e: Race is a significant predictor of whether a student graduates or not in STEM programs Data Collection The beginning data set was a 26,191 row excel file (F1) containing the transcript first semester of Fall 2003 to the Spring of 2010 Each row of the file contained demographic information on the student, the course information, the course grade information and course division information. Although the file contained a program major code there was not a grouping code that allowed com parisons between STEM and Non STEM program majors. A second data file(F2) was obtained from the study for all majors at the study college. A third data file(F3) was cre ated with the additional column of Program Group. This column contained 1 if the program major was a STEM program and 0 if the program major was Non STEM. Data files F1 and F3 were imported into Microsoft Access. Because each row in F1 only contained th e information for one course in one semester a new table(T1) was created from a query that query was run against F1, F3 and T1 to create a new data table that containe d the student demographic information, program major and course completion ratio.

PAGE 44

44 To address the research question number two, the original excel data file was imported into Microsoft Access and a query was run to create a cohort grade point average data r ecord for each student. This record contained the student demographic information and totals for the credit hours, grade points and course completion ratio. The resulting file was then imported into SPSS. Preparing a SPSS data set for research question t hree required the most preparation of data manipulation. Step One was to import the original data file into Microsoft Access and to run a query to produce a record set that contained the semester course completion ratio for any student who took either ENC 1101 College Composition or MAT1103 Intermediate Algebra. These two courses were used as gatekeepers for the purpose of this study because of the high student repeat rate. Step Two: A unique student semester identifier was created for each student in the record set. Step Three: A query was run against the original data file to create a record set that contained the semester course completion ration for any student who did not take either ENC1101 or MAT1103. A unique student semester identifier was creat ed for each student in the record set. To accomplish the actions just described actually required three intermediate steps but the Structured Query Language code is given in the appendix. Step Four: The two files were combined into one record set. The record set was then imported into SPSS in order to run a Within Subject analysis. For research question four an additional data file that contained student graduation information from Fall 2003 to Spring 2010 was obtained from the Office of Institutional R esearch. This file was imported into Microsoft Access and a query was run to create a data set that cross referenced the student identification numbers of

PAGE 45

45 students who graduated with the students in the original data cohort. This record set was then impo rted into SPSS for multiple logistic regression analysis. Data Analysis The design for this study used two quantitative methods. Longitudinal transcript analysis was used to identify barriers to program progression by showing relationships between demogra phic factors and the course completion ratio. Multiple logistic regression was used to predict STEM degree graduation. The focus of this study is to find predictors of degree progression and degree completion among STEM students. The outcome of the degre e completion question is a binary variable, either the student did or did not complete the degree. The appropriate regression approach in which the outcome variable is binary (yes/no) is logistic regression (Hosmer & Lemeshow, 2000). Descriptive and infer ential statistics were used to examine and analyze relationships between variables identified in the study. Regression modeling was used to compute correlations on all applicable variables. An alpha level of .05 was used for all statistical analysis. The statistical program SPSS was used for all statistical calculations. Table 3 1. Demographic comparisons Demographic Study Site State of Florida National Educational Sites 7 180 1,173 Female 52.4% 60% 56% Minority 35% 41% 40% Full Time 46.3% 38% 40% Pa rt Time 54.7% 62% 60% Average Age 19 25 28

PAGE 46

46 CHAPTER 4 RESULTS This study investigated the influence of several transcript variables on community college course completion ratios ( CCR) and on student graduation. The study focused on two groups of studen ts, those who have majors in the Science, Technology and Math (STEM) and those students whose majors do not include STEM programs (Non STEM). The study used several different models in order to provide the most relevant results to the research questions. This chapter provides both data on the study sample and the study findings to address the four research questions listed below. 1. Is there a significant relationship between selected student demographics (i.e. race, gender, age) and course completion ratios between program majors? 2. Is there a significant relation ship between college grade point average program major and CCR. 3. Is there a significant relationship between student enrollment patterns and CCR. 4. Is there a significant relationship between elected st udent demographics (i.e. rac e, gender, age, STEM, CCR) and likeli hood to Graduate? Descriptive Data Race The performance of different racial groups in STEM programs is the subject of many research studies. But the vast majority of these studies have bee n conducted on four year institutions. In this study of community college students Black students in STEM programs comprised 19% of the sample and 13% of the sample in Non STEM programs. Hispanic students in STEM programs comprised 10% of the sample and 9% of the sample in Non STEM. Gender Although current literature speaks to large gap in the number of female and male students in STEM programs the study demographics did not reflect this. In

PAGE 47

47 both STEM and Non STEM programs Males were 55% of the sample and females were 45% of the sample. Age This study was conducted on cohort data of first time in college students who began in Fall of 2003. As such the age factor was heavily concentrated around the 18 22 range (M=18.92, SD=3.18). STEM Group In align ment with enrollment patterns at comprehensive community colleges nationally, student enrollment was comprised of the majority of students in Non STEM programs 76% with 24% in STEM programs in the study sample. In the study sample students who declared a STEM major were coded as the value 1 in the variable STEM Group and 0 for Non STEM students. Grade Point Average Grade point average (GPA) is described in the literature as a primary indicator to college degree completion. In this study GPA was calculat ed students were found to have a statistically significantly higher GPA than Non STEM students. Although significance was found the mean GPA for STEM was barely above passing. Placement in Gatekeeper course Progression toward degree completion is prep aredness for the next higher level of course. These gatekeeper courses differ at different institutions but concentrate around math and college level reading and writing. In this study ENC1101 College Composition and MAT1033 Intermediate Algebra were cho sen for their high level of repeat attempts by students at the study college.

PAGE 48

48 Representativeness The descriptive data given above portray a student profile that is not completely inline with community college student demographics nationwide. Current resea rch data based on the National Profile of Communit y Colleges:Trends & Statistics ( Phillippe & Patton, 2000.) describe the typical community college student possessing the following demographics: Forty six percent are 25 or older, and 32 percent are at leas t 30 years old. The average age is 29. Fifty eight percent are women. Twenty nine percent have annual household incomes less than $20,000. Eighty five percent balance studies with full time or part time work. More than half (54 percent) have full time jobs Thirty percent of those who work full time also attend classes full time (12 or more credit hours). Among students 30 39 years old, the rate climbs to 41 percent. Minority students constitute 30 percent of community college enrollments nationally, with L atino students representing the fastest growing racial/ethnic population. One explanation for this departure from national norms is the close proximity of the study college with a large university in the same city. Many students attend the community colle ge in hopes of being able to transfer to the university. The community college also enrolls large numbers of students who were not admitted to the university on the first attempt. Use of Models Different statistical procedures were used to answer the rese arch questions. This approach was used to obtain the most relevant results to a particular question rather

PAGE 49

49 Research Questions Question One Is there a significant relationship between sel ected student demographics (i.e. race, gender, age) and course completion ratios between program majors? Hypothesis 1 a Male students will have higher CCR percentages than female students in STEM programs. Table 4 2 shows the results of this analysis. An independent samples t test was conducted to compare the course completion ratios for male students ( N=169, M=.609, SD=.312) and females students (N=138, M=.634, SD=.278). There was not a statistically significant difference between male and female student s [ (t(305)= .723, p=.235]. For non STEM students the difference between males (N=538, M=.572, SD=.309) and females(N=447, M=.599, SD=.324) was also found to be statistically not significant [ t(983)= 1.340, p=.09). Hypothesis 1.b Students who begin enroll ment at older ages will have higher CCR than younger peers in STEM programs Tables 4 3 through 4 7 show the result of this analysis. An significant relationship be tween age enrolled (N=307, M=18.83, SD=2.89) and course completion ratios (N=307, M=.62, SD=.30) for STEM students [r(305)=.049, p=.196]. For non STEM students there was no statistically significant relationship between age enrolled (N=985, M=18.95, SD=3.2 7) and course completion ratios (N=985, M=. 584, SD=.316). the analysis produced results of [r(983)= .049, p=.06].

PAGE 50

50 Hypothesis 1.c Black students will have lower percentages in CCR in STEM programs than all other student peer groups. Tables 4 8 through 4 11 show the result of this analysis. The results of a one way analysis of variance indicated that there were significant differences between the course completion rations of the student groups [F(5,301)=4.37, p=.001]. Tukey multiple comparisons performed a t a .05 significance level found that Black students had statistically significant lower course completion ratios than three of the five student peer groups. The mean difference, MD, and p values are as follows; Black Asian (MD= .3513, p=.001), Black Hisp anic (MD= .1866, p .045), Black White (MD= .1346, p=.023). The hypothesis was proven to be true for three of the five student groups. The hypothesis was not supported for Black Indian and Black Other comparisons. For non STEM student groups a one way ana lysis of variance indicated that there were significant differences between the course completion rations of the student groups [F(5,979)=4.62, p=.000]. Tukey multiple comparisons performed at a .05 significance level found that Black students had statist ically significant lower course completion ratios than three of the five student peer groups. The mean difference, MD, and p values are as follows; Black Asian (MD= .2316, p=.017), Black Hispanic (MD= .1467, p .007), Black White(MD= .1214, p=.001). Quest ion Two Is there a significant relationship between college grade point average, program major and course completion ratio

PAGE 51

51 Hypothesis 2.a Students in STEM majors will have a higher grade point average and course completion ratio when compared to Non STEM students. Tables 4 12 through 4 15 show the results of this analysis. An independent samples t test was conducted to compare the grade point average for STEM students (N=307, M=2.04, SD=1.03) and Non STEM students (N=985, M=1.87, SD=1.06). There was a st atistically significant difference between STEM and Non STEM students [ t(1290)= 2.47, p=.006]. To compare course completion ratios an independent samples t test was conducted to compare the course completion ratios for STEM students (N=307, M=.621, SD=.30 ) and Non STEM students (N=985, M=.584, SD=.3160). There was a statistically significant difference between STEM and Non STEM students [ t(1290)= 1.80, p=.04]. The hypothesis was supported for STEM students demonstrating a higher grade point average and a higher course completion ratio than Non STEM students. Hypothesis 2.b STEM students with higher grade point averages will demonstrate higher course completion ratios Tables 4 15 throu gh 4 20 show the result of this analysis. An correlation coefficien t indicated that there is a statistically significant relationship between grade point average (N=307, M=2.04, SD=1.03) and course completion ratios (N=307, M=.62, SD=.30) for STEM students [r(305)=.968, p=.000]. The hypothesis was supported by the results of the analysis. For Non STEM students the strong relationship was also evident [r(985)=.966, p=.000).

PAGE 52

52 Question Three Is there a significant relationship between student enrollment patterns and course completion ratios? Hypothesis 3. a Higher credit hour enrollment positively impacts cours e completion ratios. Tables 4 21 4 23 show the result of this analysis. Term enrollment was found to be a statistically significant predictor of course completion ratios (b= .004, p= .000). A one credit hour increase in term enrollment will decrease course completion ratio by .004. This result does not support the hypothesis and seemingly counters literature stating that higher credit hour enrollment increases engagement and therefore increase s course completion rates. However this may be due to the high mean of credit hour term enrollment (M=10.79, SD=4.80). Given that full time enrollment is defined by most institutions as 12 credit hours, the results of this analysis point to the negative effect of enrolling past 12 credit hours. Hypothesis 3. b Enrollment in either a gatekeeper course or an academic foundation course negatively impacts course completion ratio. Table 4 24 4 25 show the result of this analysis. To simplify the analysis on ly two courses were chosen as study variables. For gatekeeper status enrollment in ENC1101 was used. For academic foundation status MAT1033 was used. The courses were chosen because they have one of the highest repeat rates at the study college. Enroll ment in either of these courses was coded as a 1 in the Gatekeeper variable. The gatekeeper variable was not shown to have a statistically significant relationship to the course completion ratio (b=.003, p= .746). The hypothesis was not supported by the results of the analysis.

PAGE 53

53 Question Four Is there a significant relationship between selected student demographics (i.e. race, gender, age, STEM, CCR) and likelihood to graduate? Multiple logistic regression analysis was used to predict graduation from two year programs using age at enrollment, gender, course completion ratio, and race. The analysis was conducted for STEM and Non STEM students in order to compare results. Hypothesis 4.a Female students enrolled in STEM programs have higher odds of graduati ng than male peers Table 4 29 shows the result of this analysis. Gender was found to be statistically significant predictor for graduation for STEM students (b= 1.125, z=.573, p=.001). For students in STEM programs the odds ratio for Female was .3 25, indicating that the odds of graduating for females is .325 of that of males, a decrease of 68%. These results do not support the hypothesis and are contrary to current research showing that females graduate at a higher rate than males across academic programs. However, the results do support literature that states the environment for females and minorities in STEM programs is not conducive to healthy, progressive academic environment (Tietjen, 2004). In examining the non STEM cohort gender was not fo und to be statistically significant (b= .319, z=1.81, p=.070). Hypothesis 4.b Older students are more likely to graduate than younger students in STEM programs. Table 4 29 shows the result of this analysis. Age enrolled was not found to be a statistic ally significant predictor for graduation when controlling for all other variables for STEM students (b= .007, z=. 19 p=. 847 ). However this result could be

PAGE 54

54 influenced by the student sample being heavily clustered around the 18 20 age group. A more even ly distributed sample may have produced different results. This hypothesis was not supported by the analysis. In examining non STEM students Age enrolled was found to be statistically significant (b= .110, z=2.81, p=.005). Hypothesis 4.c Course completion ratio is a significant predictor of whether a student graduates or not in STEM programs. Table 4 30 shows the result of this analysis. CCR was found to be statistically significant predictor when controlling for all other variables for STEM students(b=.0 80, z=8.73, p=.000). The Odds Ratio for the statistic was very large (Exp(b)=3129) because it refers to the effect that a CCR change from 0 to 1 has on the odds of graduating. To provide a more easily understood assessment of the dependence of graduation on CCR, the odds ratio for a .10 change in CCR was computed. This odds ratio was 2.24 indicating that for students enrolled in STEM programs, the odds of graduating increase by 224% when CCR changes by .10. The hypothesis was supported by the analysis. F or non STEM students the odds ratio for a .10 change in CCR was computed. CCR was found to be statistically significant predictor when controlling for all other variables (b=.663, z=14.64, p=.000). The odds ratio was 1.94 indicating that for students enr olled in non STEM programs, the odds of graduating increase by 194% when CCR changes by .10. Hypothesis 4.d Being enrolled in a STEM program is a significant predictor of graduating. Table 4 31 shows the result of this analysis. To test the hypothesis a multiple logistical

PAGE 55

55 regression analysis was run using all cases with enrollment in a STEM program being coded as 1 or 0. The variables Race, Age at enrollment and CCR were also included in the analysis. Enrollment in a STEM program was not found to be statistically significant (b= .218, z=1.27, p=.204). This hypothesis was not supported by the analysis. Hypothesis 4.e Race is a significant predictor of whether a student graduates or not in STEM programs. Table 4 29 shows the result of this analysis. STEM programs was not found to be statistically significant (z=1.73,p=.703). This hypothesis was not supported by the analysis. In non STEM programs race was not found to be statistically significant (z=2.09, p=.496). Summary This chapter provided the analysis and findings associated with the four research questions posed by this dissertation study. This study used both descriptive and inferential statistical analyses to provide results. Each hypothesis was found either to be supp orted or not supported through examination of the results. Several approaches were used to answer the research questions including Pearson Correlations, Multiple Logistical Regression and Univariate Analysis of Variance and Independent Samples t tests.

PAGE 56

56 Ta ble 4 1. Group Statistics Group Gender Group N Mean Std. Deviation Std. Error Mean Non STEM CCR Male 538 .572 .3087 .0133 Female 447 .599 .3243 .0153 STEM CCR Male 169 .609 .3123 .0240 Female 138 .634 .2782 .0237

PAGE 57

57 Table 4 2 Independent Sample s Test Group Levene's Test for Equality of Variances t test for Equality of Means 95% Confidence Interval of the Difference F Sig. t df Sig. (2 tailed) Mean Difference Std. Error Difference Lower Upper Non STEM CCR Equal variances assumed 2.315 .128 1.340 983 .180 .0271 .0202 .0668 .0126 Equal variances not assumed 1.334 931.608 .182 .0271 .0203 .0669 .0128 STEM CCR Equal variances assumed 2.196 .139 .723 305 .470 .0247 .0341 .0918 .0425 Equal variances not assumed .732 302.677 465 .0247 .0337 .0911 .0417

PAGE 58

58 Table 4 3 Descriptive Statistics Mean Std. Deviation N Age at enrollment 18.95 3.271 985 CCR .584 .3160 985 a. Group = Non STEM Table 4 4 Correlations Age at enrollment CCR Age at enrollment Pearson Correlatio n 1 .049 Sig. (2 tailed) .121 N 985 985 CCR Pearson Correlation .049 1 Sig. (2 tailed) .121 N 985 985 a. Group = Non STEM

PAGE 59

59 Table 4 5 Descriptive Statistics Mean Std. Deviation N Age at enrollment 18.83 2.889 307 CCR .621 .2973 307 a. G roup = STEM Table 4 6 Correlations Age at enrollment CCR Age at enrollment Pearson Correlation 1 .049 Sig. (2 tailed) .393 N 307 307 CCR Pearson Correlation .049 1 Sig. (2 tailed) .393 N 307 307 a. Group = STEM

PAGE 60

60 Table 4 7 Descriptiv es CCR N Mean Std. Deviation Std. Error 95% Confidence Interval for Mean Minimum Maximum Lower Bound Upper Bound Asian 22 .705 .2989 .0637 .572 .837 .1 1.0 Black 134 .473 .2850 .0246 .425 .522 .0 1.0 Hispanic 93 .620 .2948 .0306 .559 .681 .0 1 .0 Indian 5 .557 .3974 .1777 .064 1.050 .0 1.0 Other 18 .657 .3367 .0794 .490 .825 .0 1.0 White 713 .595 .3197 .0120 .571 .618 .0 1.0 Total 985 .584 .3160 .0101 .564 .604 .0 1.0 a. Group = Non STEM

PAGE 61

61 Table 4 8 Non STEM ANOVA CCR Sum of Square s df Mean Square F Sig. Between Groups 2.265 5 .453 4.620 .000 Within Groups 95.990 979 .098 Total 98.255 984 a. Group = Non STEM Table 4 9 Mean Comparisons CCR Tukey HSD (I) Code for Race (J) Code for Race Mean Difference (I J) Std. Error Si g. 95% Confidence Interval Lower Bound Upper Bound Black Asian .2316 .0720 .017 .437 .026 Hispanic .1467 .0423 .007 .267 .026 Indian .0835 .1426 .992 .491 .324 Other .1837 .0786 .180 .408 .041 White .1214 .0295 .001 .206 .037 a. Group = Non STEM

PAGE 62

62 Table 4 10 Descriptive STEM groups CCR N Mean Std. Deviation Std. Error 95% Confidence Interval for Mean Minimum Maximum Lower Bound Upper Bound Asian 13 .847 .2041 .0566 .724 .970 .4 1.0 Black 59 .496 .2680 .0349 .426 .566 .0 1.0 Hispanic 31 .682 .2760 .0496 .581 .783 .0 1.0 Indian 2 .848 .1357 .0959 .371 2.067 .8 .9 Other 5 .650 .2913 .1303 .288 1.012 .2 1.0 White 197 .630 .3022 .0215 .588 .673 .0 1.0 Total 307 .621 .2973 .0170 .587 .654 .0 1.0 a. Group = STEM

PAGE 63

63 Table 4 11 ANOVA STEM group CCR Sum of Squares df Mean Square F Sig. Between Groups 1.831 5 .366 4.373 .001 Within Groups 25.207 301 .084 Total 27.037 306 a. Group = STEM Table 4 12. STEM comparisons Multiple Comparisons CCR Tukey HSD (I) Code for Race (J) Code for Race Mean Difference (I J) Std. Error Sig. 95% Confidence Interval Lower Bound Upper Bound Black Asian .3513 .0887 .001 .606 .097 Hispanic .1866 .0642 .045 .371 .002 Indian .3520 .2081 .538 .949 .245 Other .1542 .1348 .863 .541 .232 White .1346 .0429 .023 .258 .011 *. The mean difference is significant at the 0.05 level. a. Group = STEM

PAGE 64

64 Table 4 13 T test Group Statistics Group N Mean Std. Deviation Std. Error Mean G.P.A. Non STEM 985 1.876 1.0595 .0338 STEM 307 2.047 1.0283 .0587 Table 4 14 GPA T test Levene's Test for Equality of Variances t test for Equality of Means 95% Confidence Interval of the Difference F Sig. t df Sig. (2 tailed) Mean Difference Std. Error Diff erence Lower Upper G.P.A Equal variances assumed .317 .574 2.474 1290 .013 .1702 .0688 .3051 .0352 Equal variances not assumed 2.513 524.17 7 .012 .1702 .0677 .3032 .0372

PAGE 65

65 Table 4 15 Course Completion Ratio Group Statistics Group N Mean Std. Deviation Std. Error Mean CCR Non STEM 985 .584 .3160 .0101 STEM 307 .621 .2973 .0170 Table 4 16 Course Completion Ratio T test Levene's Test for Equality of Variances t test for Equality of Means 95% Confidence Interval of the Differenc e F Sig. t df Sig. (2 tailed) Mean Difference Std. Error Difference Lower Upper CCR Equal variances assumed 2.499 .114 1.797 1290 .073 .0366 .0204 .0766 .0034 Equal variances not assumed 1.855 538.73 9 .064 .0366 .0197 .0754 .0022

PAGE 66

66 Table 4 17 Non STEM descriptives Mean Std. Deviation N CCR .584 .3160 985 G.P.A. 1.876 1.0595 985 a. Group = Non STEM Table 4 18 Non STEM Correlations CCR G.P.A. CCR Pearson Correlation 1 .966 ** Sig. (2 tailed) .000 N 985 985 G.P.A. Pearson C orrelation .966 ** 1 Sig. (2 tailed) .000 N 985 985 **. Correlation is significant at the 0.01 level (2 tailed).

PAGE 67

67 Table 4 19 STEM Descriptives Mean Std. Deviation N CCR .621 .2973 307 G.P.A. 2.047 1.0283 307 Table 4 20 STEM Correlations CCR G.P.A. CCR Pearson Correlation 1 .968 ** Sig. (2 tailed) .000 N 307 307 G.P.A. Pearson Correlation .968 ** 1 Sig. (2 tailed) .000 N 307 307 **. Correlation is significant at the 0.01 level (2 tailed).

PAGE 68

68 Table 4 21 Descriptive Statistic s N Minimum Maximum Mean Std. Deviation TermEnr 7090 .1 37.0 10.786 4.7959 Valid N (listwise) 7090 Table 4 22 Tests of Between Subjects Effects Dependent Variable:CCR Source Type III Sum of Squares df Mean Square F Sig. Corrected Model 443.157 a 1294 .342 3.425 .000 Intercept 329.399 1 329.399 3294.444 .000 TermEnr 2.131 1 2.131 21.314 .000 StdID 441.825 1293 .342 3.418 .000 Error 579.419 5795 .100 Total 4224.707 7090 Corrected Total 1022.577 7089 a. R Squared = .433 (Adjusted R S quared = .307)

PAGE 69

69 Table 4 23. Parameter Estimates. Dependant Variable CCR 95% Confidence Interval Parameter B Std. Error t Sig. Lower Bound Upper Bound Intercept .694 .158 4.383 .000 .384 1.005 TermEnr .004 .001 4.617 .000 .006 .002 Table 4 24 Tests of Between Subjects Effects Dependent Variable: CCR Source Type III Sum of Squares df Mean Square F Sig. Corrected Model 441.037 a 1294 .341 3.396 .000 Intercept 1019.678 1 1019.678 10161.010 .000 Gatekeeper .011 1 .011 .105 .746 StdID 433 .004 1293 .335 3.337 .000 Error 581.540 5795 .100 Total 4224.707 7090 Corrected Total 1022.577 7089 a. R Squared = .431 (Adjusted R Squared = .304)

PAGE 70

70 Table 4 25. Parameter Estimates. Dependant Variable CCR 95% Confidence Interval Parame ter B Std. Error t Sig. Lower Bound Upper Bound Intercept .646 .158 4.007 .000 .335 .956 TermEnr .003 .009 .324 .746 .015 .021 Table 4 26. Dependent Variable Encoding BegSTEM Group Original Value Internal Value Non STEM NON Graduate 0 Graduate 1 STEM NON Graduate 0 Graduate 1

PAGE 71

71 Table 4 27. Case Processing Summary Multiple Logistic Regression Results BegSTEM Group Unweighted Cases a N Percent Non STEM Selected Cases Included in Analysis 1001 100.0 Missing Cases 0 .0 Total 1001 100.0 Unselected Cases 0 .0 Total 1001 100.0 STEM Selected Cases Included in Analysis 324 100.0 Missing Cases 0 .0 Total 324 100.0 Unselected Cases 0 .0 Total 324 100.0

PAGE 72

72 Table 4 28. Categorical Variables Codings Multiple Logistic Regression Results BegSTEMGroup Frequency Parameter coding (1) (2) (3) (4) (5) Non STEM Race A 22 1.000 .000 .000 .000 .000 B 134 .000 1.000 .000 .000 .000 H 94 .000 .000 1.000 .000 .000 I 5 .000 .000 .000 1.000 .000 O 18 .000 .000 .000 .000 1.000 W 728 .000 .000 .000 .000 .000 Gender F 453 1.000 M 548 .000 STEM Race A 14 1.000 .000 .000 .000 .000 B 60 .000 1.000 .000 .000 .000 H 34 .000 .000 1.000 .000 .000 I 3 .000 .000 .000 1.000 .000 O 5 .000 .000 .000 .000 1.000 W 208 .000 .000 .000 .000 .000 Gender F 145 1.000 M 179 .000

PAGE 73

73 Table 4 29. Variables in the Equation Multiple Logistic Regression R esults BegSTEMGroup B S.E. Wald df Sig. Exp(B) Non STEM Step 1 a Female .319 .176 3.291 1 .070 .727 Race 4.383 5 .496 Race(1) .204 .571 .127 1 .721 1.226 Race(2) .466 .300 2.411 1 .120 .628 Race(3) .118 .286 .170 1 .680 .889 Race(4) 2 1.026 14949.663 .000 1 .999 .000 Race(5) .815 .669 1.488 1 .223 2.260 CCR 6.628 .453 214.342 1 .000 756.006 AgeEnr .110 .039 7.944 1 .005 .896 Constant 2.940 .786 13.981 1 .000 .053 STEM Step 1 a Female 1.125 .329 11.680 1 .001 .325 Race 2.981 5 .703 Race(1) .288 .755 .145 1 .703 1.333 Race(2) .504 .452 1.248 1 .264 1.656 Race(3) .296 .471 .395 1 .529 .744 Race(4) 20.683 23148.049 .000 1 .999 9.602E8 Race(5) 1.251 1.237 1.021 1 .312 3.492 CCR 8.049 .921 76.379 1 .000 3129.616 AgeEnr .007 .039 .037 1 .847 .993 Constant 5.522 .974 32.139 1 .000 .004 a. Variable(s) entered on step 1: Gender, Race, CCR, AgeEnr.

PAGE 74

74 Table 4 30. CCR multiplied by 10 Multiple Logistic Regression Results BegSTEMGroup B S.E. Wald df Si g. Exp(B) Non STEM Step 1a Female .319 .176 3.291 1 .070 .727 Race 4.383 5 .496 Race(1) .204 .571 .127 1 .721 1.226 Race(2) .466 .300 2.411 1 .120 .628 Race(3) .118 .286 .170 1 .680 .889 Race(4) 21.026 14949.663 .000 1 .999 .000 R ace(5) .815 .669 1.488 1 .223 2.260 CCRx10 .663 .045 214.342 1 .000 1.940 AgeEnr .110 .039 7.944 1 .005 .896 Constant 2.940 .786 13.981 1 .000 .053 STEM Step 1a Female 1.125 .329 11.680 1 .001 .325 Race 2.981 5 .703 Race(1) .288 .755 .145 1 .703 1.333 Race(2) .504 .452 1.248 1 .264 1.656 Race(3) .296 .471 .395 1 .529 .744 Race(4) 20.683 23148.049 .000 1 .999 9.602E8 Race(5) 1.251 1.237 1.021 1 .312 3.492 CCRx10 .805 .092 76.379 1 .000 2.236 AgeEnr .007 .039 .037 1 847 .993 Constant 5.522 .974 32.139 1 .000 .004 a. Variable(s) entered on step 1: Gender, Race, CCRx10, AgeEnr.

PAGE 75

75 Table 4 31. All Cases Multiple Logistic Regression Results B S.E. Wald df Sig. Exp(B) Step 1 a Female .514 .153 11.296 1 .001 .598 Race 4.162 5 .526 Race(1) .228 .446 .262 1 .609 1.256 Race(2) .203 .242 .709 1 .400 .816 Race(3) .160 .245 .425 1 .514 .852 Race(4) .389 .883 .193 1 .660 .678 Race(5) .920 .594 2.397 1 .122 2.509 AgeEnr .067 .025 7.123 1 .008 .935 CCRx 100 .069 .004 292.238 1 .000 1.072 BegSTEMGroup(1) .218 .172 1.612 1 .204 .804 Constant 3.691 .564 42.779 1 .000 .025 a. Variable(s) entered on step 1: Gender, Race, AgeEnr, CCRx100, BegSTEMGroup.

PAGE 76

76 Table 4.32. Classification BegSTEMGroup Obs erved Predicted GradStatus Percentage Correct NON Graduate Graduate Non STEM Step 1 GradStatus NON Graduate 557 101 84.7 Graduate 94 249 72.6 Overall Percentage 80.5 STEM Step 1 GradStatus NON Graduate 160 32 83.3 Graduate 29 103 78.0 Overall Percentage 81.2 a. The cut value is .500

PAGE 77

77 Figure 4 1 Plot of Means, Non STEM groups

PAGE 78

78 Figure 4 2 Plot of Means STEM groups

PAGE 79

79 Figure 4 3 Box Plot of GPA

PAGE 80

80 Figure 4 4. Histogram of age groupings of students in study

PAGE 81

81 CHAPTER 5 DI SCUSSION Purpose The purpose of this dissertation was to indentify barriers to the progression to degree completion of the Information Technology student of first time in college community college students. The Information Technology degree belongs to the group of majors in the science, technology, math and engineering collation that is often referred to as STEM. Institutional data at the study college shows that STEM students often change their degree choice while at the community college level. In orde r to capture the best representation of barriers to degree progress this study followed the student cohort of all STEM majors. In this study several quantitative methods were used for analysis to give the best result of the pertinent research question. T he course completion ratio was used as a valid construct to measure the progress to degree. The use of the course completion ratio is grounded in the research conducted by the Transfer and Retention of Urban Community College Students (TRUCCS) project, wh ich used the course completion ratio as a measure of progress to degree completion of community college students. This study presents the findings of the research questions using both STEM students and Non STEM students in order to investigate if a differe nce exists between the two groups in the dependant variables being measured. This study used both demographic and enrollment related variables. The variables chosen were included base on a review of the literature that indentifies major factors that inf luence degree completion. Student demographic variables included age, race and gender. Enrollment

PAGE 82

82 average, program major and course completion ratio. By using transcr ipt analysis this study was able to translate enrollment data into new coded variables such as enrollment in gatekeeper course and course completion ratio when enrolled in gatekeeper course. Multiple logistic regression was used to predict odds to graduat ion for several student groups including STEM and Non STEM. In addition, logistic regression analysis allowed for results to be presented on how membership in certain demographic categories such as age or race influenced odds to graduate. The relationshi p of the findings and the four research questions are presented in this chapter along with discussions of the limitations of this study, implications of findings and suggestions for future research. Comparison with Previous R esearch This section will compa re the use of the course completion ratio in a past study, Kress(2007), to the findings of this study to see if the course completion ratio resulted in the same data indicators. In comparing study results with the research of Kress(2007) not all of the sa me independent variables were tested as in this study so only instances where the variables are the same are the comparisons made. It should be noted that the Kress(2007) study did not divide students into program major groups but in this study students w ere either in STEM or Non STEM groups. Comparison 1 Difference in Male and Female Course Completion Ratios Kress(2007) found a significant difference in the course completion ratios of male and female students. Kress(2007) findings state for male students (M=.581, SD=.309) and those for female students (M=.635, SD=.290). There was a statistically significant difference at the p=.001 level in course completion ratios for male and female students [t(1381)= 3.385, p=.001]. While there significance was found the actual mean

PAGE 83

83 difference between scores was small. In this current study no significant difference between male and female students course completion ratios were found in both the STEM and Non STEM student groups. Comparison 2 Student Age and the Course Completion R atio The Kress(2007) study and this current study both found there was no statistically significant relationship between student age and the course completion ratio. Comparison 3 Race and the Course Completion R atio The Kress(2007) study and t he current study both found that African American students had the lowest course completion ratio of all student racial groups. These findings agree with the literature base that states African American students are performing at the lower end of the scal e on academic predictors of success. Comparison 4 Grade Point Average and the Course Completion R atio The Kress(2007) study and this current study both found that grade point average and the course completion ratio were highly correlated. In the Kress(200 7) study grade point average was removed from the final regression model over concerns that it may confound other results. In this study the correlation between grade point average and the course completion ratio was [r(305)=.968, p=.000] for STEM student s and [r(985)=.966, p=.000) for Non STEM. Discussions of Findings Course Completion Ratio This study used the course completion ratio as a valid construct to measure progress toward degree completion. The course completion ratio is defined as the number o

PAGE 84

84 assumption grade as admissible for progression it is not a universal academic practice. Interpretation of the Course Completion Ratio The interpretation of the course completion ratio can be influenced by several factors. In using the course completion ratio several decisions must be made before analyses can begin. First, does the transcript data account for gr ade forgiven ess? Grade forgiveness is the policy of allowing the latest grade in a repeated course to be the only grade recorded on the transcript. The policy effects the number of hours attempted by the student and changes the calculated course completi on ratio. In this study the transcript only records the last grade given, the original course attempt is not recorded. Second, does the transcript data include grades of incomplete or withdrawals? This also will affect the total number of hours attempted In this study the course completion ratio does include hours for incomplete or withdrawal grades. Third, the course completion ratio is highly correlated with grade point average. Therefore the course completion ratio is best used as an indicator of momentum while the student is still enrolled once each semester. Used in this manner the course completion ratio is an progression toward degree. The average course comp letion ratio for all students in the study was 59%. For female students in STEM programs the mean course completion ratio was 63% and for male STEM students the mean course completion ratio was 61%. For Non STEM females the course completion ratio was 60 % and form male Non STEM students the ratio was 57%. These findings suggest that the two year associates

PAGE 85

85 degree is two year in name only and that most students will take much longer than two years to finish the degree. One important issue that rises from this result is that 40% or more of the average student semester course load is not completed. Given the research the research in the literature that speaks to the variety of negative outcomes associated with withdrawals or failures (Windham, 2006; Zhao, 1999) it is conceivable to believe that if the study was extended it would be found that many of the students would never graduate. This result agrees with the findings of studies by Adelman(2004) that showed less than 50% of the community college student cohorts finish within six years. Even for those students who finish the degree the path is a long one. The average enrollment of semester credit hours was 11. Given that the average course completion ratio was 60% that gives a progression to degree of 6 .5 successful hours per semester. The Associate of Arts degree is 60 credit hours. This means the average student will need approximately nine semesters to finish the degree. If the student could attend year round that would take three years to complete The previous example is a best case scenario that assumes the student is not starting in remedial course work and is not a multiple repeater in prerequisite sequence. For the STEM student this is especially troubling given that most of the math and sci ence course curriculum is delivered in sequence. For example Physics One must be completed before Physics Two, Calculus One must be successfully completed before Calculus Two. There are several core subjects that have this sequential course listing. The question of funding under these circumstances becomes complex. Most students optimistically believe and plan for their two year degree to take close to two years to complete with a cushion of

PAGE 86

86 perhaps one semester. Therefore as they plan financially they structure their finances accordingly. The reality of the actual length of time causes many students to encounter financial difficulties and cause them to abort their college goals. The course completion ratio also provides insight in terms of enrollment management. Taking into account the data that shows only 60% of courses taken will be finished successfully the college has an opportunity to increase seat counts knowing there will be attrition as the semester progress. To actualize this procedure would take substantial planning to predict which courses will have the attrition and at what time in the semester. This finding also has implications for higher education in the realm of academic advisement. The students who are in the 40% that do not finish coursework successfully need to have more intense academic advisement. The question that should be explored is why 40% of courses do not complete successfully It seems that students somehow are either self advising or being given incorrect advisement. If the 40% of unsuccessful courses can be realigned to become completions the affected college should realize a financial increase do the recaptured tution. Question One This research question explored the relationship between student demographics and the course completion ratio for the STEM and Non STEM students. The que stion sought to examine if there existed innate differences in the biological differences in student groups (age, race, gender) and the students performance in course completion ratios. Hy pothesis 1.a : Male students will have higher CCR percentages than female students in STEM programs The study found no significant difference between male and female students in their course completion ratios in STEM and in Non STEM

PAGE 87

87 student program majors This finding is interesting considering literature (Goennner & Snaith, 2004) that state female students out perform their male peers in enrollment and in course progression to degree. This finding also diverges from literature (Tietjen, 2004) that repo rt the widening gap between male and female students in the STEM programs. This finding suggests an opportunity for further research to investigate whether this cohort was unique in some way that caused it to diverge from current assumed norms in the lite rature. One possible approach to investigate this divergence would be to conduct interview with student focus groups to explore differences in how males and females interact and approach their studies in the STEM programs. One possible answer is that fem ale students do not have large enough numbers to form sustaining support and study groups that would enhance their academic performance. Hypothesis 1.b: Students who begin enrollment at older ages will have higher CCR than younger peers in STEM programs Although in general terms an increase in age corresponds to an increase in maturity which in theory should lead to the ability to focus on academic course work, increased age is generally found to be in negative correlation to degree completion (Gao, 2002) The rational for this negative relationship becomes clear when the non academic circumstances of the older student are unveiled. Issues such as job requirements, family responsibilities and financial obligations are sometimes overwhelming for the older student (McHewitt, 1993). However, the results of this study found no statistically relationship between age and course completion ratios. This result could be attributed to the several factors. One this study was limited to first time in college stude nts which skewed the data toward traditional entry age of

PAGE 88

88 college freshman. Nearly 83% of study sample contained student ages between 18 and 20 years. Hypothesis 1.c: Black students will have lower percentages in CCR in STEM programs than all other studen t peer groups Many studies exist suggesting that a Blau 1999; Astin, 2001). The reasons for this relationship are complex and actually involves the combination of severa l factors including financial, environmental and academic preparation (Bailey, 2004; Pascarella & Terenzini, 2005). In this study Black students were shown to have statistically lower course completion ratios than White, Hispanic and Asian students. Howe ver, as supported by the literature the reasons for the difference in performance can not be attributed solely to race. An obvious opportunity exists for further research to reveal additional variables in the performance indicators of Black students. Ques tion Two This research question explored the relationship between enrollment data such as program major and grade point average and the course completion ratio. The question rose out of a review of the literature that identifies a positive relationship be tween grade point average and course completion ratios (Gao, 2002; Gebel 1995). While it is intuitive to assume that higher performing students, those with higher grade point averages, will have higher completion rates a student can have a completion rate of 100% with only a grade point average of 2.0. The questions also explores whether STEM students have completion ratios on par with Non STEM students. Hypothesis 2.a: Students in STEM majors will have a higher grade point average and course completion r atio when compared to Non STEM students. STEM students

PAGE 89

89 were found to have statistically significantly higher grade point averages and course completion ratios than Non STEM students. The reason for this difference could fall into several categories. STE M students begin their programs with the knowledge that the course work is perceived to be more challenging than general liberal arts studies and therefore may approach their course work in a more focused manner. Another explanation is the sequential natu re of the STEM courses. Because many of the core subjects require a two or three course sequence that builds on the subject matter, STEM students may subject indoctrination factor that gives them a frame work of context as they progress through the course sequence. While it is true that there are course sequence examples in Non STEM course work, it is not as detailed or prevalent as in STEM curriculum. Concern is raised for both STEM and Non STEM students when the actual grade point averages and course co mpletion ratios are studied. STEM students mean grade point average was only 2.04 and their course completion ratio was 62%. For Non STEM students the grade point average was only 1.87 and the course completion ratio was 59%. If STEM students are more m otivated than Non STEM students it is only by a small measurable difference. Hypothesis 2.b: STEM students with higher grade point averages will demonstrate higher course completion ratios There was a very strong positive relationship between grade point average and course completion. This result is supported by the literature in several studies (Nippert, 2000; Windham, 1994). An interesting aspect of this result points to the elimination of course repeats which accelerates time to degree completion. W hile this result again is intuitive it also points

PAGE 90

90 is possible for a student with a grade point average of 2.0 to have a course completion ratio of 100%. In this sample this high completing average student was not found. This may because a grade point average of 2.0 is an indicator of a student that is on the verge slipping into a stop out status (Pascarella & Terenzini, 2005). Question Three This question explore d the relationship between student enrollment patterns and course completion ratios. A review of the literature states part time enrollment has a negative influence on college course completion. In other words if a student is enrolled for less than full time hours the course completion ratio decreases. Hypothesis 3 a: Higher credit hour enrollment positively impacts course completion ratios A review of the literature states that part time enrollment has a negative influence on course completion ratios ( Adelman, 2006; Bailey, 2004). The results of this study report that there was a negative relationship to enrollment hours and course completion ratio. In other words, as the enrollment increases the course completion ratio decreased. On first observatio n this results seems to contradict the findings of the literature but on closer examination a reasonable explanation can be proposed. The mean semester enrollment was 10.8 hours. Given that full time enrollment is defined as 12 hours students in the samp le were close to full time. When it is taken into account that the average course is three credit hours then additional course beyond the average would place the student in more than full time status. Therefore the results suggest that course enrollment beyond 12 hours is detrimental to course completion ratios. This explanation is sensible given the academic work load that a full time student is under. When the additional factors such as family obligations,

PAGE 91

91 work requirements and financial stress are ad ded in it is easy to see why a student would have declining academic success beyond 12 hours per semester. It is also noted that this analysis was conducted as a within person statistical approach that did not control for other variables and that the actu al effect while significant was very small considering the sample size. Hypothesis 3 b: Enrollment in either a gatekeeper course or an academic foundation course negatively impacts course completion ratio A review of the literature finds several studies that state the relationship between academic unpreparedness and completion rates (Floyd, 2002; Gao, 2002; Gerardi, 1996; Hoyt, 1999; Lajubutu & Yang). However, these studies focused on what is defined as remedial or below college level course work. While these are of course important issues this study examined the effect of college level high repeat courses referred to in some studies as gatekeeper courses. These are courses that are college level but students attempt them multiple times before successfu lly passing. At the study college ENC1101 and MAT1033 have are reported by the office on Institutional Research as having one the highest repeat repeats. This study chose those variables because both STEM and Non STEM students would have to successfully pass them in order to complete their degree programs. This study found no statistically significant relationship between being registered for a high repeat course and the course completion ratio. While the results of this study seem counter intuitive, th ere exists several possible scenarios that may provide an answer. One explanation maybe the nature of the gatekeeper course itself. Perhaps the method in which these courses are taught is so different than the norm at the study college that failure to co mplete the course has not an indication of student

PAGE 92

92 ability but rather of inadequate course design. Another explanation maybe the sample used in this study did not reflect the typical experience of students at the study college. Question Four This question explored the relationship between student demographic variables such as age, gender, race along with enrollment variables such as program group and course completion ratio to the like hood of graduating. The literature states that demographics such as ra ce and gender due have a relationship to completion rates and graduation ( Peltier, Laden, & Matranga, 1999; Wassmer, Moore, & Shulock, 2004). This study question sought to investigate if this relationship existed in the sample of community college student s and if there existed a difference between STEM and Non STEM students in terms of student demographics and student graduation. Hypothesis 4.a: Female students enrolled in STEM programs have higher odds of graduating than male peers Current research sta tes that women out perform their male peers in completion and graduation rates (Bailey, Calcagno, Jenkins, Kienzl, & Leinbach, 2005; Goenner & Snaith, 2004). However in the STEM programs women are still not to the level of male peers in numbers in the pro gram majors or in completion rates (Tietjen, 2004). This study found that women graduated at a rate 68% lower than males in STEM programs. It is unlikely that genetics is the cause for the difference in performance. Rather the results point to an alarmi ng statistic that obligates future research to be performed on the root cause of this discrepancy. One probable explanation is because of the low number of female STEM students a critical mass has not yet formed that allow for a self sustaining student su pport group. Without a student network of peer support female students may feel more isolated and less likely to seek academic help.

PAGE 93

93 Hypothesis 4.b : Older students are more likely to graduate than younger students in STEM programs. While research states lower than younger students ( Gutierrez Marquez, 1994), this study found age not to be statistically significant predicator of graduation. The hypothesis was based on the assumption that the focus required to compl ete course work in STEM programs would be more prevalent in older students. While the study found age not to be a predictor of graduation this result could have been influenced by the highly skewed sample. More than 80% of the students in the sample were between the ages of 18 and 21. Hypothesis 4.c : Course completion ratio is a significant predictor of whether a student graduates or not in STEM programs Graduation is the result of continuous positive completion rates. So it is intuitive that course co mpletion rates to be a predictor of graduation. But this question sought to explore how strong that relationship is. It is possible for a student to have course completion ratio of below 50% and still graduate. Actually in theory as long as the course c ompletion ratio is above zero as student would eventually graduate assuming the student is not repeating the same courses. Course completion ratio was found to be a very significant predictor of whether a student graduates or not. In analyzing the effect of the course completion ratio this study transformed the ratio from ranging in value from 0 to 1 to ranging in value from 1 to 10. This allowed an easier interpretation of the odds ratio. It was found that a 0 .1 increase in the course completion ratio for STEM students increased the odds of graduating by 224%. This significant increase indicates that graduating is not simply the completion of credit hours but it is the manner in which the hours are completed that is important. For

PAGE 94

94 Non STEM students a similar result was found. A 0 .1 increase in course completion ratio increased the odds of graduation by 194% in Non STEM students. Hypothesis 4.d : Being enrolled in a STEM program is a significant predictor of graduating This question explored whether a student enrolled in a STEM program graduated at higher odds than Non STEM students. This question arose from the assumption that STEM students knowing they face a challenging curriculum are more committed to graduation and therefore have higher graduatio n odds. The results reported that there was no significant relationship between program major group (STEM vs. Non STEM). This result may be influenced by the very low graduation rates of both program groups. The graduation rate of STEM students was only 34% and for Non STEM students the rate was 41%. While it may seem that program major does not influence graduation odds, the more pressing issue seems to be the low rate of graduation regardless of major. Hypothesis 4.e : Race is a significant predictor o f whether a student graduates or not in STEM programs. The literature is uniform in the belief that race in general is a predictor for course completion and graduation. In particular Black and Latino students are at risk for low performance and low grad uation rates ( Bailey, Calcagno, Jenkins, Kienzl, & Leinbach, 2005; Peltier, Laden, & Matranga, 1999). This question explored whether this phenomenon would be present in the study sample. In this study race was not found to be a significant predictor of g raduation odds. This contradiction with the literature may be explained in the low graduation rate of the study sample. This result may be in accord with results at other community college where graduation rates continue to be below 50% (Adelman, 2006). Because the graduation rate for all

PAGE 95

95 students at the study site was below 45% the influence of race on graduation may be more difficult to detect. If graduation rate was higher it is believed the influence of race on graduation would be more evident. Limi tations This study was limited by the following factors: Study Sample The study sample consisted of first time in college students at one community college. This study used Associate of Arts seeking students and did not include Associate of Science seek ing students. The sample age was skewed in the 18 21 age range and the results may not generalize to the broader community college student demographics. Geographic Location The study college was located in Northeast Florida in the same city as a large p their studies at the community college and then transfer to the university. As such, the sample may include traditional demographics of a n entering freshman at a univer sity rather than the demographics of a typical community college student. Implications of the Findings In terms of demographic data this study found several results that seem to be contradictory with published literature. Male students were found to be co mpleting at ratios on par with their female peers and age was not found to significant predictor of course completion. These results however must be taken in the light that the average course completion ratio was only 60%. The study results certainly imp ly an opportunity to raise completion rates among all demographic groups. One area that the literature

PAGE 96

96 states and the study results supported were the alarming completion rates of Black students. Black students performed lower than White, Hispanic and As ian student groups in STEM and Non STEM groups. Intervention methods should be created to close the performance gap between Black student and other racial groups. Given the limited funding all community colleges face, target funding at the lowest perform ing groups should yield the highest return on investment dollars. STEM students were found to have higher completion ratios and grade point averages than Non STEM students. There are several implications concerning this result. While an argument could be made that STEM students have higher performance rates because they are more dedicated to a curriculum known to be rigorous, the countering argument could be that Non STEM students should have higher performance rates because the Non STEM curriculum is les s rigorous. It seems more plausible that the difference in performance lies in the difference in which the curriculum is taught. Colleges have an opportunity to study logical sequencing and pedagogy of STEM programs and adapt the methods to Non STEM area s. For example the use of gaming simulations can be a excellent teaching aide in a variety of subject areas from English Composition to Art History. The effect of over enrollment was borne out in this study. While part time enrollment is a negative influ ence on degree completion, over enrollment is also counterproductive Colleges should strive to couns el students to strike a balance with enrollment while striving to stay as close to the 12 credit limit as possible. Over and under enrollment can be done on occasion but neither approach is beneficial in repeated terms.

PAGE 97

97 Graduation is the ultimate goal of a student in degree program and graduation should be, and is, the focus of most colleges and universities. This study revealed that females in the STEM p rograms at the study college graduated at a rate 68% lower than male students. This should be an unacceptable rate at any college. The implication of this finding is stretches into the nature of the study problem, the lack of students in the Information Technology degree. If female students continue to gain on male students in enrollments then the success of female students in the program will affect the overall success of the degree. The retention of female students in the STEM programs should be an in stitutional priority at every college. Recommendations for Further Research 1. As a quantitative study this research did n ot use interviews or other qualit ative methods to cross reference findings. An opportunity exists to conduct interview within the studen t groups to document their story and provide reasons to the reported outcomes. 2. The performance of Black students continues to be an area that colleges struggle with. This study adds to the literature base that more research must be done to find the reason s for the performance gap. 3. This study focused on one community college located next to a large university. Another opportunity for research exists in expanding the study to several community colleges located in various settings to compare results on perfo rmance. 4. The underperformance of females in the STEM programs reported by this study gives an opportunity for further research to disclose if this statistic is being observed at other colleges. 5. The low graduation rates of all students demographics in the st udy requires further study. The reasons may very well be complex but an opportunity for a qualitative more evidence to solve this issue. 6. The role of the gatekeeper course has additional aspects to be explored. Questions such as the effect of multiple repeats vs only one repeat on persistence to graduation may offer insights on which courses require curriculum redesign.

PAGE 98

98 7. The higher performance of STEM students suggests that there lies an unnamed variable that accounts for the performance gap. This variable may be qualitative or it may lie in the course design differences of the STEM programs.

PAGE 99

99 LIST OF REFERENCES Adelman, C., & Educational Resources Information Center (U.S. ). (1990). Light and shadows on college athletes microform : College transcripts and labor market history Washington, DC: U.S. Dept. of Education, Office of Educational Research and Improvement, Educational Resources Information Center. Adelman, C., & Ed ucational Resources Information Center (U.S.). (1992). The way we are microform : The community college as american thermometer Washington, DC: U.S. Dept. of Education, Office of Educational Research and Improvement, Educational Resources Information Cent er. Adelman, C., & Educational Resources Information Center (U.S.). (1997). Leading, concurrent, or lagging? microform : The knowledge content of computer science in higher education and the labor market Washington, DC: U.S. Dept. of Education, Office of Educational Research and Improvement, Educational Resources Information Center. Adelman, C., & Educational Resources Information Center (U.S.). (2000). A parallel postsecondary universe [microform] : The certification system in information technology Wa shington, DC: U.S. Dept. of Education, Office of Educational Research and Improvement, Educational Resources Information Center. Adelman, C., & Institute of Education Sciences (U.S.). (2004). Principal indicators of student academic histories in postsecon dary education, 1972 2000 [electronic resource] Washington, D.C.: Institute of Education Sciences, U.S. Dept. of Education. Adelman, C., & United States Dept. of Education. (1998). Women and men of the engineering path : A model for analyses of undergrad uate careers Washington, DC: U.S. Dept. of Education; National Institute for Science Education. Adelman, C., National Science Foundation (U.S.), United States Dept. of Education, & University of Wisconsin Madison National Institute for Science Education. (1997). Leading, concurrent or lagging? : The knowledge content of computer science in higher education and the labor market Washington, DC; Madison, WI: United States Department of Education; National Institute for Science Education. Akbulut, A. Y., & Looney, C. A. (2007). Inspiring students to pursue computing degrees. Communications of the ACM, 50 (10), 67 71. Allen, W. R., (Ed.), & And Others. (1991). College in black and white: African american students in predominantly white and in historically bla ck public universities. state university of new york series, frontiers in education

PAGE 100

100 Anthes, G. (2006). Computer science looks for A remake. Computerworld, 40 (18), 26 28. Arora, S., & Chazelle, B. (2005). Is the thrill gone? Communications of the ACM, 48 ( 8), 31 33. Ayala, C., & Striplen, A. (2002). A career introduction model for first generation college freshmen students Babco, E. L., Zumeta, W., & Raveling, J. (2000). Is the nation's top talent opting out of science and engineering? Issues in Science & Technology, 17 (1), 28. Bandura, A. (1993). Perceived self efficacy in cognitive development and functioning. Educational Psychologist, 28 (2), 117. Calcagno, J., Crosta, P., Bailey, T., & Jenkins, D. (2007). Stepping stones to a degree: The impact of enr ollment pathways and milestones on community college student outcomes. Research in Higher Education, 48(7), 775 801. Camp, A. G., Gilleland, D., Pearson, C., & Vander Putten, J. (2009). Women's path into science and engineering majors: A structural equati on model. Educational Research & Evaluation, 15(1), 63 77. Campbell, N. J. (1992). Enrollment in computer courses by college students: Computer proficiency, attitudes, and.. Journal of Research on Computing in Education, 25 (1), 61. Ching Huei Chen. (2007 ). Cultural diversity in instructional design for technology based education. British Journal of Educational Technology, 38(6), 1113 1116. Compeau, D. R., & Higgins, C. A. (1995). Computer self efficacy: Development of a measure and initial test. MIS Quart erly, 19 (2), 189 211. Denning, P. J., & McGettrick, A. (2005). Recentering computer science. Communications of the ACM, 48 (11), 15 19. Forte, A., & Guzdial, M. (2005). Motivation and nonmajors in computer science: Identifying discrete audiences for intro ductory courses. IEEE Transactions on Education, 48 (2), 248 253. Foster, A. L. (2005). Student interest in computer science plummets. Chronicle of Higher Education, 51 (38; 38), A31 A32. Fountain, J. E. (2004). Searching for the keys to unlock the clubhou se. IEEE Technology & Society Magazine, 23 (2), 6 9.

PAGE 101

101 Gao, H. (2002, October). Examining the length of time to completion at a community college. Paper presented at the Annual Meeting of the Southern Association for Institutional Research, Baton Rouge, LA. (ERIC Document Reproduction Service No. ED475 987). Garrison, D. R., Cleveland Innes, M., Koole, M., & Kappelman, J. (2006). Revisiting methodological issues in transcript analysis: Negotiated coding and reliability. The Internet and Higher Education, 9 (1 ), 1 8. Gebel, M. (1995, May). Impacts on baccalaureate degree completion: A longitudinal analysis of community college transfer students. Paper presented at the Annual Meeting of the Association for Institutional Research, Boston, MA. (ERIC Document Repr oduction Service No. ED387 002). Gibson, S. (2005). IT work force gap looming. EWeek, 22 (48), 37 37. Goldrick Rab, Sara. (2007). Promoting Academic Momentum at Community Colleges: Challenges and Opportunities. CommunityCollege Research Center Working Pape r #5. Teachers College, Columbia University. Retrieved August 2010 from http://ccrc. Goldstein, M. T., & Perin, D. (2008). Predicting performance in a community college area course from academic skill level. Community College Review, 36(2), 89 115. Grange r, M. J., Dick, G., Jacobson, B. M., & Van Slyke, C. (2007). Information systems enrollments: Challenges and strategies. Journal of Information Systems Education, 18 (3), 303 311. Hagedorn, L. S., & Kress, A. M. (2008). Using transcripts in analyses: Direc tions and opportunities. New Directions for Community Colleges, 2008(143), 7 17. Hayes, F., & Speaking, F. (2004). Offshore dot bomb. Computerworld, 38 (44), 52 52. Holahan, C. (2007). The myth of high tech outsourcing. Business Week Online, 29 29. Hol mes, N. (2007). The computing profession and higher education. Computer, 40 (1), 116 115. Horn, L., Adelman, C., Chen, X., United States Dept. of Education, & United States Office of Educational Research and Improvement. (1998). Toward resiliency : At risk students who make it to college Washington, DC: U.S. Dept. of Education, Office of Educational Research and Improvement. Hosmer, D. & Lemeshow, S. (2000). Applied Logistic Regression. Chicago, Wiley Interscience Publication

PAGE 102

102 Increasing College Completion Retrieved November 4, 2010, from www.gatesfoundation.org Information technology education enrollment Ishitani, T. T. (2006). Studying attrition and degree completion behavior among first generation college students in the united states. Journal of High er Education, 77(5), 861 885. Jaeger, A. J., & Eagan, J.,M.Kevin. (2009). Examining the effect of part time faculty members on associate's degree completion. Community College Review, 36(3), 167 194. Kevin Eagan, M., & Jaeger, A. (2009). Effects of expos ure to part time faculty on community college transfer. Research in Higher Education, 50(2), 168 188. Koefoed, J. O., Jr. (1984). An evaluation of the use of major selection to predict program completion rates of career oriented liberal arts programs at k irkwood community college. Nova University). Report: ED256422. 38p. (Level 1 Available online, if indexed January 1993 onward) Kolhede, E. (2001). Gender effects on the major selection process -A five year study: Implications for marketing business pr ograms of small private Col1eges to women. Journal of Marketing for Higher Education, 11 (2), 39 60. Lerman, R. I., Riegg, S. K., & Salzman, H. (2000). The role of community colleges in expanding the supply of information technology workers Lerman, R. I., Riegg, S. K., & Salzman, H. (2001). Community colleges: Trainers or retrainers of workers. Community College Journal, 71 (6), 41 44. Lightfield, E. T., & Rice, D. D. (1974). Course sequence length as a determinant of academic major selection. Lillibridge F. (2008). Retention tracking using institutional data. New Directions for Community Colleges, 2008(143), 19 30. McBride, N. (2007). Erase old programme and launch new version McCormick, A. C., Carroll, C. D., & National Center for Education Statistics (1999). Credit production and progress toward the bachelor's degree [microform] : An analysis of postsecondary transcripts for beginning students at 4 year institutions Washington, D.C.: U.S. Dept. of Education, Office of Educational Research and Improv ement; For sale by the U.S. G.P.O., Supt. of Docs. Murphy, C. (2005). SPEAK UP for the IT career. InformationWeek, (1058), 34 41.

PAGE 103

103 National Center for Education Statistics. (2005). First generation students in postsecondary education [electronic resource] : A look at their college transcripts -postsecondary education descriptive analysis report Washington, D.C.: United States Dept. of Education, National Center for Education Statistics. Ntiri, D. W. (2001). Access to higher education for nontraditional s tudents and minorities in a technology focused society. Urban Education, 36 (1), 129. Perrakis, A. I. (2008). Factors promoting academic success among african american and white male community college students. New Directions for Community Colleges, 2008(1 42), 15 23. Projected job openings 2006 2016 U.S. Bureau of Labor Statistics. Soh, L., Samal, A., & Nugent, G. (2007). An integrated framework for improved computer science education: Strategies, implementations, and results. Computer Science Education, 17 (1), 59 83. Targeted industry sectors Workforce Florida, Inc. The College Completion Agenda. Retrieved November 4, 2010. http://completionagenda.collegeboard.org/reports Tietjen, J. S. (2004). Why so, few women, still? IEEE Spectrum, 41 (10), 57 58. T ietjen, J. S. (2004). Why so, few women, still? IEEE Spectrum, 41(10), 57 58. Tucker, A. B. (1999). Enrollments and staffing in college computer science programs: A Growth perspective for 1996 2000. Computer Science Education, 9 (1), 023. Vegso, J (2008). Enrollments and Degree Production at US CS Departments Drop Further in 2006 07. Computing Research News, 20 (2). Wang, X. (2009). Baccalaureate attainment and college persistence of community college transfer students at four year institutions. Research i n Higher Education, 50(6), 570 588. Ward, B. (2007). Computer science enrollments drop in 2006. Computer, 40 (6), 85 85. Whalen, D. F., & Shelley II, M. C. (2010). Academic success for STEM and non STEM majors. Journal of STEM Education: Innovations & Res earch, 11(1), 45 60.

PAGE 104

104 BIOGRAPHICAL SKETCH Eugene Garrison Jones II was born in Ft. Myer s Florida. He holds a Bachelor of Science degree in computer science and a Master of Education in education leadership both from the University of Florida. He has he ld positions as tenured Associate P rofessor of Computer Science, Coordinator of Computer Science and Director of Information Technology Education at Santa Fe Col lege. He has authored several capitalization i mprovement grants to enhance technology classroo ms and was the director of a state appropriation to design and implement the delivery of computers, internet access and training to 200 hundred families in Northeast Florida. Under his leadership the Interactive Media Department was recognized as an Innov ation of the Year winner by the League of Innovation in the Community College. He has presented at several national conferences on using technology to enhance and improve the learning environment and the administration of the college department. Mr. Jone s also reaccreditation requirement by the Southern Association of Colleges and Schools.