Historically one of the factors used to determine admittance to graduate schools has been the Graduate Record Examination (GRE). However, is GRE performance a predictor of success on the Standard Principal Texas Examination of Educator Standards (TExES)? Prior to receiving educator certification in Texas, a candidate must secure a satisfactory score on the TExES as established by the State Board for Educator Certification (SBEC). SBEC is the entity legislatively charged with overseeing all certification of educators in the State of Texas. The philosophy implied in this system is that certification assessments help to ensure that successful candidates possess the proficiencies necessary for success in the particular position. A recent study of 278 students conducted by Newby and Littleton (2001) established that passing the Standard Principalship TExES could be predicted by using a student’s GRE verbal and analytical subtest scores and their ethnicity. Since that time the GRE has discontinued administering the analytical subtest. No other literature on predicting performance on the Standard Principal TExES was identified.

Chapter 241 of the Texas Administrative Code (2001) establishes the requirements for a Standard Principal certification which is required for an administrative position as an assistant principal or principal in a Texas public school. Effective September 1, 2000, prior to admission to an administrator preparation program, an individual must: 1) hold a baccalaureate degree from an accredited institution of higher education; and, 2) demonstrate an acceptable combination of a score on a nationally-normed assessment and grade point average determined by the preparation program. Most universities comply with the second item by requiring the GRE for admission to their administrator preparation program. Principal certification in Texas requires the certification candidate to hold a masters degree; complete a state approved administrator preparation program; and, earn a passing score on the Standard Principal TExES. The Standard Principal TExES content tests knowledge, skills and abilities in nine competencies clustered around three domains (see Appendix).

The Standard Principal TExES was developed by soliciting input from a cross section of K-12 educators, faculty from educator preparation programs, education service center personnel, professional organizations, content experts, business community and parents. The TExES development process included the following phases:

1. Developing standards for what beginning principals should know;

2.Reviewing standards through public comment period;

3.Developing the framework which included the specific competencies;

4.Conducting a content validation survey to determine the relative job importance of each competency;

5. Developing and reviewing test item by the test developer;

6.Conducting a pilot test of the new items;

7.Reviewing the pilot test data and conducting preliminary standards setting;

8. Administering the new TExES Principal test;

9. Setting the passing standards (State Board for Educator Certification, 2004).

Graduate Record Examination (GRE)

Historically, one of the criteria used for screening and admitting a student to a graduate school is the use of the Graduate Record Examination (GRE) developed by the Educational Testing Service (2005). The premise underlying the test suggests a relationship between a particular score attained on the GRE and the likelihood of future success in graduate level courses.

The Graduate Record Examination’s General Test measures an individual’s verbal, quantitative, and analytical skills that have been acquired over time and are not related to a specific field of study. The verbal component measures ability to analyze and evaluate material and synthesize information obtained from it. The quantitative measures basic mathematical skills and the understanding of elementary mathematical concepts (Educational Testing Service, 2005). The analytical writing score was implemented in October 2002 and designed to measure the individuals’ ability to articulate complex ideas clearly and effectively and sustain a well-focused and coherent argument. The analytical writing score was not used in this research study.

Each subtest yields a score between 200 – 800. Scores from the different subtests should be considered independent of one another because each measure is scaled separately. For instance, based on all examinees who took the test between October 1, 1996 and September 30, 1999 a score of 600 on each of the subtests would place the individual at the 85th percentile in verbal, 55th percentile in quantitative and 61st percentile in analytical. During the same period, the mean scores of all test takers were 471 verbal, 569 quantitative and 547 analytical. However, the mean scores for examinees who reported education as their major were 447 verbal, 511 quantitative and 534 analytical (Educational Testing Service, 2004).

According to the Educational Testing Service (2005), the Graduate Records Examination scores provide a common measure for comparing the abilities of students who come from a variety of colleges and universities with different standards and methods for measuring success in their program. However, it should be noted that the Educational Testing Service does not advocate using the GRE in isolation but as a supplement to other information such as undergraduate record, letters of recommendation, interviews, or other information used in the admissions process.

Educational Testing Service (2004) conducted a study during the 1993-1994 testing year (n = 417K) to determine the effect on different subgroups when a combined verbal and quantitative score of 1000 was used for admission to graduate school. For instance 86% of African American, 69% Mexican American, 61% of Hispanics, whereas, only 45% of whites and 38% of Asian Americans fell below the combined verbal and quantitative score of 1000. With regard to gender, 55% of females scored below the cut off as opposed to 35% of males. The study concluded that the use of adhering to a particular cut off score could systematically discount large numbers of students from various subgroups. Interestingly, Davidson and Lewis (2000) studied the predictive validity of GRE, IQ and general academic aptitude but failed to establish any correlation between GRE and criterion-referenced examinations.

The design for this study was correlational research. Students enrolled in a Standard Principal certificate program at two regional universities served as the source of data. The study involved collection of two test scores from each participant in an effort to analyze any potential correlation and/or predictive value of one to the other.

Most accredited universities and graduate programs require some form of standardized test, such as the Graduate Records Examination (GRE), as criteria for program admission. The two universities involved in this study use the Verbal and Quantitative scores on the GRE as part of their respective admission criteria. Given this fact, the research in this study was guided by the question: What is the value of the GRE in predicting performance on the Standard Principal Examination for Certification of Educators (TExES) in Texas?

The scores from two instruments were used in this study. First, scores from the verbal and quantitative portions of the Graduate Records Examination were combined (referred to as Combined Verbal/Quantitative GRE Score) to make one score for each student in the study. Possible scores on each portion of the Graduate Records Examination range from 200 to 800 (Educational Testing Service, 2005). The maximum score that could be earned on the combination of verbal and quantitative was 1600.

Additionally, the total score earned by each student on the Standard Principal Examination for the Certification of Teachers (TExES) was also used. The term TexES replaced the previously used term ExCET, however the change in name did not affect the examination itself, therefore the use of scores when the examination was called ExCET is equally appropriate as with the new name of TExES.

In fall 2000, new Principal and Superintendent ExCET tests were introduced. Although these two new tests represented the first stage of the new testing and certification program, the new program had not yet been renamed as TExES. As of

fall 2002, these two tests have been designated as TExES tests to reflect this change; however, the content and structure of these tests have not changed ( State Board for Educator Certification, Front Cover).

This study utilized the total scale score earned on the examination and did not utilize the individual domain scores. Thus, the possible scores on this instrument ranged from 100 to 300.

The sample for this study consisted of students enrolled in two regional university administrator preparation programs. The group from one university (University A) was composed of all students taking the Standard Principal TExES during the first Spring 2002 test administration date. The group from the other university (University B) in the study was composed of all students taking the Standard Principal TExES during the second Spring 2002 test administration date. There were no significant differences in the two testing periods.

Thirty students from the University A and 23 students from University B collectively created the sample for this study of 53 students. Of the 53 students making up the sample, 54.7% (n = 29) were female and 45.3% (n = 24) were male. Ethnically, the study sample divided in to three sub-populations. White students, with 73.6% (n = 39) of the sample, constituted the largest ethnic group. African-American students accounted for 17% (n = 9), followed by Hispanic students at 9.4% (n = 5).

Combined Verbal/Quantitative GRE Scores for the students enrolled at University A showed a low score of 510 and a high score of 1080. University A score analysis documented a mean of 796.33 for the Combined Verbal/Quantitative GRE Scores with a standard deviation of 161.79. University B Combined Verbal/Quantitative GRE scores showed a range with a minimum score of 580 and a maximum score of 1170. The mean for University B scores was 854.78 with a standard deviation of 143.68.

Standard Principal TExES scores between the two University groups were somewhat similar. University A produced a TExES score range of a minimum of 195 and a maximum of 284. The mean for the University A TExES scores was 249.70 with a standard deviation of 20.80. University B TExES scores produced a minimum score of 217 and a maximum score of 279. The mean of the University B scores was 253.22 with a standard deviation of 14.43.

A correlation between the two scores was determined by utilizing the Pearson Product-Moment Coefficient (r). Statistical analysis of the sample data of Combined Verbal/Quantitative GRE Scores and Standard Principal TExES scores provided r scores of 0.686, thus indicating the existence of a correlation between the GRE and TExES scores. Statistical significance of this correlation was calculated at the p = 0.01. With the correlation established the scores were plotted on a scatterplot in order to statistically draw a line to determine the linear regression (see Figure 1). The regression line provided a visual representation of slope relative to the swarm of data and demonstrated the best prediction of Y from X for any given data-point.

The least squares regression line passes through the data according to the derived linear equation [y = 0.0768x + 187.92] whereas y is the predicted TExES score and x represents the actual Combined Verbal/Quantitative GRE score. The standard error of slope was found to be 0.011289.

In order to test the null hypothesis that the true slope is zero, the ratio of the obtained slope to its standard error was calculated. This calculation yielded a t-test statistic of [0.0768 / 0.011289 = 6.803] showing statistical significance (p < 0.00001). Thus, the coefficient for the GRE was significantly different from zero, and this provided substantial evidence that scores on TExES Number 68, Standard Principal, were significantly influenced by GRE performance. Additionally, the coefficient of determination (r2) was calculated to be 0.4711. The coefficient of determination indicated that about 47% of the variation in TExES score performance was explained by performance in the Graduate Records Examination thus supporting the use of the GRE as a predictor on the Standard Principal TExES. However, the coefficient of determination was not useful as a causative calculation.

Finally, the standard error of estimate, to indicate an index of prediction error, was determined to be 13.294. Given this value, an individual who obtained a mean combined GRE score of 1000 is predicted to score 265 on the Standard Principal TExES. Moreover, 95% of the time, such a student will score between 238 and 291 on a similar test. Interestingly, this student has a 96.57% probability of obtaining a passing score (240) on the Standard Principal TExES.

A student who scores only 800 on the Combined Verbal/Quantitative GRE is predicted to score 249 on the TExES. In this scenario, the 95% confidence interval concludes a Standard Principal TExES score between 223 and 276, and the student has a 75.82% probability of passing the examination. The aggregated data displayed on Figure 1 seem to suggest that the probabilities for passing the Standard Principal TExES decrease substantially when a student scores below 800 on the Combined Verbal/Quantitative Graduate Records Examination.

The data document the significant relationship and correlation of the Graduate Records Examination (GRE) and the Standard Principal Examination for the Certification of Teachers in Texas (TExES) within the sample population. With such a relationship and correlation established, the regression line derived from the data offered an important and helpful equation in predicting a Standard Principal TExES score from a GRE score. This supported the use of the GRE to meet the state requirement that a standardized test be used in admission and screening standards for certification candidates. This also may assist university faculty, involved in Standard Principalship graduate certification programs, in the screening and admission process.

While the sample included students from two different Texas public institutions of higher education, the total number of students in the study (n=53) should be expanded. Replication of this study incorporating data from additional Universities would be useful. Additionally, similar studies in other States using the GRE for admission to administrator preparation programs would also provide valuable data on this discussion. Finally, as the revised Standard Principalship TExES continues to be administered, additional data at the two Universities represented in the study and others will become more abundant thus offering additional opportunity for study.

The shortage of qualified candidates to assume the role of principal in the future must be addressed. With states moving toward performance-based accountability models that require administrative candidates to pass a state licensing examination, it becomes incumbent upon administrator preparation programs to attract individuals who have the potential to pass the requisite licensing examination. One approach is to determine the predictive validity of the Graduate Record Examination to the state-licensing exam. Of course the critical question that remains: What correlation exists between TExES and actual successful performance on the job?