SAT

<<Back to the EDC 529 home page

Is there a correlation between Scholastic Assessment Test (SAT) scores and completion of graduation requirements for women of different ethnicities?
 * Research Question**

//Research hypothesis (non-directional)//: There will be a correlation between SAT scores and completion of graduation requirements between women of different ethnicities.
 * Hypotheses**

//Null hypothesis//: There will be no correlation between SAT scores and completion of graduation requirements between women of different ethnicities.

When students are admitted into a college or university, predicting whether or not they will graduate is a difficult problem. Standardized tests, such as the SAT, are often used in college admissions to predict freshmen success in college. The College Board (the administrators of the SAT) states that “validity studies consistently find that high-school grades and SAT scores together are substantial and significant predictors of achievement in college” (Camara and Echternacht, 2000). However, most of these studies focus on freshman grade point average and not graduation rates, although they are correlated (Stanley and Abe, 1988; Geiser and Studley, 2002; Kim et al., 2003).
 * Literature Review**

There have been many studies on regional levels of the correlation between SAT scores and student graduation rates within that region or school (York et al., 1993; Waugh et al., 1994; St. John et al., 1999; Kim et al., 2003). Most of these studies, with the exception of the study done by Waugh and others (1994) show that college admissions test scores (SAT) appear to "consistently be significant predictors of retention" (Astin, Koren, and Green as cited by Reason, 2003). According to a study done by Tross, SAT scores accounted for "29% of the variance in retention" (Reason, 2003). However, according to Bangura (1995) only family income is statistically and significantly correlated with SAT scores, with children of high income families scoring higher on the SAT than other students. One of the problems with doing a study of a particular school or group of schools is that if a student transfers to another university, they are not counted as graduating, even if they were successful. In addition, student behavior in individual schools or states may not be representative of students nationally.

The studies that have been completed show that students of different ethnicities graduated at different rates and men and women graduated at different rates (York et al., 1993; Kim et al., 2003). African-American, Hispanic, and Native American students were less likely than White or Asian-American students to graduate (Kim et al., 2003; Murtaugh, as cited by Reason, 2003). Geiser and Studley (2002) state that students who are minorities are predicted to have a higher first-year GPA than they actually do. Gender issues were more complicated, with some studies reporting that women were more likely to graduate (York et al., 1993; Kim et al., 2003; Eric Clearinghouse on Urban Education, 2001; Pelitie as cited by Reason, 2003). "The increase for females over the past 25 years was very substantial, almost 12 percentage points, whereas the increase for males was only 2 percent" (Eric Clearinghouse on Urban Education, 2001). Some studies report that men were more persistent (St. John et al., 1999), and some studies report that there was no gender difference (Stanley and Abe, 1988). However it has been noted that there is a "substantial race and ethnic gap, with the completion rate for whites at 28 percent; blacks, 16 percent; and Hispanics, 10 percent." (Eric Clearinghouse on Urban Education, 2001). According to McEvans and Astin (as cited by Pavel, 1992) only "15 percent of the Native Americans earned BS degrees, compared to 37% of Whites and Asians". Both Male and Female Natives were unlikely to be retained (Astin, as cited by Pavel, 1992).

In addition to SAT scores and the effects of gender and ethnicity, other factors play a strong role in influencing whether or not students graduation from college. It has also been suggested by Leafgran (as cited by Pritchard and Wilson, 2003) that students who are emotionally and socially healthy have a greater chance to succeed in college and that many psychological variables impact College GPA and retention.

This study will examine graduation rates for women of different ethnicities. The ethnicity of the women will be determined by what they self reported on their college admission form. To determine the graduation rate, we will determine which women earned an undergraduate degree within a 6-year time span from when they entered college. Previous studies have used 5 years as a cut-off for graduation (York et al., 1993; Kim et al., 2003). However, students enrolled in college programs often take 6 years to complete their degree and must be included in this undergraduate population.

Our research showed a slight correlation between SAT scores and completion of graduation requirements. Larger sample sizes are needed to confirm these results. However, we recognize there are other factors which contribute to a student's success. While higher education may be more aware of factors which contribute to retention, we need to take the initiative to improve graduation rates. When this is done, our expectations of success will be achieved.

We will randomly sample 50% of incoming, female freshman at Universities and Colleges in Rhode Island (i.e. Rhode Island College, Rhode Island School of Design, Providence College, University of Rhode Island, Roger Williams University, Bryant University, Salve Regina University, and Brown University). The random numbers table will be used to generate our sample and establish the list of participants. The following female ethnic groups will be used to complete our sample: White, Asian, African American, Hispanic, American Indian and other.
 * Sample**:

The data collected will be SAT scores and whether or not the student completed graduation requirements. SAT scores will be collected as part of their admissions requirement. Successful completion will be analyzed over a 6 year period. The instrument to determine graduation will be the degree audit system.
 * Data Collection**:

A correlation coefficient will be used to measure the relationship (correlation) between SAT scores and the completion of graduation requirements within six years.
 * Data Analysis:**

According to Orcher (2005), quantitative researchers favor instruments (tests) that yield consistent scores and consistency is synonymous with reliability. Our study has content validity based on experts' judgments of the appropriateness of the contents of the test, in our case the SAT. The content validity is determined by having experts compare the contents of a test with the instructional objectives contained in a curriculum guide.
 * Validity and Reliability:**

M**ortality** threatens the validity of our study – As individuals may withdraw or transfer from their University or College, not to leave school, but to finish at another institution. Since all college students may withdraw or transfer, we will address this weakness with random sampling and sample size, 50% of Rhode Island's Universities and Colleges incoming freshman.

The threat of history in this study is affected by the fact that we are studying women from diverse backgrounds. Our study will address these weaknesses by constructing a random sample of students. This random sample of students will give us (as close as we can possibly get) an equivalent population at the onset of our study and sample size, 50% of Rhode Island's Universities and Colleges incoming freshman.
 * History** also threatens the validity of our study --

Because of the design of our study, we will be including fewer non-traditional students in our study group than they are represented in the population, and minorities make up a large number of non-traditional students. Non-traditional students are more likely to attend community college first and transfer, and they are less likely to graduation within the 6-year window of our study. As a result, we will need to interpret our results keeping in mind that we have fewer non-traditional students in our sample. For the group of students that are in our study, we will have a random sample.
 * Sampling Selection Bias** also threatens the validity of our study --

And finally, **Sampling Error and Chance Error** is also a threat to our validity -- To reduce this error, we are using inferential statistics with random sampling and a large sample size (50% of the population).

Ours is a quantitative study that can be replicated and the findings can be generalized to similar groups. A weakness to this study is transfer students who attend community colleges that are not required to submit SAT scores as part of the application process. Universities have a large population of transfer students; therefore, it may be difficult to separate the two groups when analyzing rates of graduation.
 * Strength of Design**:

In addition, our study will be of students in Rhode Island, and therefore the results may not be generalizable to the population of the entire United States. The population of students in Rhode Island is not the same as the population of students throughout the United States. However, in Rhode Island, there are many schools of different levels of achievement, so our sample will cover a wide range of colleges and student abilities.

A qualitative study would be an effective way to follow up this study. Once we determine if there is a difference in the correlation between graduation rates of women of different ethnicities and their SAT scores, we would like to do a qualitative study to find out why there is a difference. We would use a semi-structured interview and purposefully select students who cover the range of variation we observed in SAT scores, graduation rates, and ethnicities.

1. What factors influenced you to stay in school and graduate or leave school? 2. Do you think your SAT score is a reflection of your academic preparedness for college? 3. Why did you go to college? Did you initially plan to graduate with a degree?
 * Examples of questions that might guide the qualitative study are:**

Bangura, A.K. (1995). Hold Everything!: Emerging Problems in Institutional Accountability for Retention and Graduation. (ERIC ED410 845)
 * References**

Camara, W. & Echternacht, G., (2000). The SAT I and high school grades: Utility in predicting success in college//.// //College Board Report No. RN-10.// New York: College Entrance Examination Board.

Eric Clearinghouse on Urban Education, (2001). Gender Differences in Educational Achievement within Racial and Ethnic Groups. (ED 455341)

Geiser, S. & Studley R., (2002). UC and the SAT: Predictive validity and differential impact of the SAT I and SAT II at the University of California. //Educational Assessment//, 8 (1) 1-26.

Kim, M.M., Rhoades, G., & Woodard, D.B., (2003). Sponsored research versus graduating students? Intervening variables and unanticipated findings in public research universities. //Research in Higher Education//, 44 (1) 51-81.

Orcher, L.T., (2005). Conducting Research, Social and Behavior Science Methods. Pyrczak Publishing, CA.

Pavel, M., (1992). American Indians and Alaska Natives in Higher Education: Research on Participation and Graduation. (ERIC EDO-RC-92-2).

Pritchard, M., & Wilson, G., (2003). Using Emotional and Social Factors to Predict Student Success. //Journal of College Student Development,// Vol 44, no 1, 18-26.

Reason, R., (2003). Student Variables that Predict Retention: Recent Research and New Developments. //NASPA Journal//, Vol 40, no. 4, 172-185.

St. John, E.P., Hu, S., Simmons, A., & Musoba, G.D., (1999). Aptitude vs merit: What matters in persistence? //Indiana// //Education Policy Center//, Policy Research Report #99-1.

Stanley, S. & Abe J. (1988). Predictors of academic achievement among Asian American and white students. //College Board Report// No. 88-11.

Waugh, G., Micceri, T., & Takalkar, P., (1994). Using ethnicity, SAT/ACT scored, and high school GPA to predict retention and graduation rates. (ERIC ED453704).

York, C.M., Bollar, S., & Schoob, C. (1993). Causes of college retention: A systems perspective. (ERIC ED369027).

The distribution of duties on this project was shared equitably. All areas were worked on collaboratively with ideas shared equally among members of the team.
 * Distribution of Duties**

Wiki technology made this collaberate assignment achieveable given the time constraints of a summer course. It was easy to use and communicate with group members. The Team Research Design gave me further insight into edcational research. I also found visiting other EDC529 reseach designs interesting.
 * Dale**

The Wiki technology was kind of a neat idea. It really did make working on the project much easier to complete as it reduced the amount of out-of-class time that is required for group work**.**
 * Cathy**

I enjoyed doing this research project. I learned alot, and I read journals that I would never had read, but glad I did. I also found learning the Database very helpful. I am sure I will use it for other classes. I particularly like the fact that we did this as a group. To do all the work alone and keep up with the rest of the demands of the course would have been very difficult for me. I feel very fortunate to have had a very good group of women who worked hard to produce a great (in my opinion) finished product. A definite plus to the course.
 * Denise**

This project was an integral part of the class because it allowed us to apply some of the concepts that we learned. It was great to be able to get feedback from other members of the group instead of working on the project on our own. The wiki allowed us to work as a group fairly easily, and luckily our group did a great job of everyone contributing and listening to everyone else's opinion. It would have been nice to know in advance when Dr. Adamy was going to look at our site to give us comments so we could be sure to have our most recent version or discussion up. Also, I think I might have learned more about experimental designs if we were forced to pick a topic that required us to do a true experiment or quasi-experiment instead of being able to do a correlation study.
 * Karen**

Denise Gorenski is a Specialist in Student Affairs in the College of Pharmacy at The University of Rhode island. She is working on her Masters Degree in Adult Education. The Research Project was part of the requirements for the completion of EDC 529.
 * Authors**

Karen Kortz is a professor at the Community College of Rhode Island and is a student in the Geology Ph.D. program at the University of Rhode Island. Her research focus is teaching geology to college-level students. This project was a requirement for the completion of EDC 529.

Cathy Tessier is a student at the University of Rhode Island in the Masters in Adult Education program. This Team Research Design was prepared to fulfill the requirements of EDC 529 Foundations of Educational Research.

Dale P. Wildenhain is a student at University of Rhode Island in the Master of Education / Teacher Certification program. This Team Research Design was compiled as a requirement of EDC 529 Foundations of Educational Research.

<<Back to the EDC 529 home page