Development of a Student Perception Inventory.

Underachievement in gifted students has been described by Gowan as "one of the greatest social wastes of our culture" (1955, p. 247). The National Commission on Excellence in Education (1983) reported that approximately half of the gifted students in the United States do not perform to the level of their tested abilities. Because gifted underachievers usually perform at satisfactory or average levels, they are often not identified as either gifted or in need of intervention programs (Statues, 1988). However, since underachievement constitutes a gap between potential and performance, these bright but underachieving students constitute a loss of productivity to themselves and to society.

There is no universally accepted definition of underachievement. Some researchers focus on standardized instruments alone to define underachievement (Supplee, 1990), while others place more emphasis on student actions in the classroom (Baum, Renzulli, & Hebert, 1995). However, most researchers agree that underachievement is a discrepancy between expected performance, based on some standardized measure of ability, and actual performance (Emerick, 1992; Peterson & Colangelo, 1996; Whitmore, 1980). Baker, Bridget, and Evans (1998) suggest that educators think broadly about the multifaceted problem of underachievement in gifted students. Their research with 26 gifted underachievers and 30 gifted achievers revealed that causes of underachievement could be traced to the family, school, and characteristics of the child. Their examination of students in one school district indicated poor compatibility between the needs and learning style of the gifted child and the classroom environment.

While the overall problem of underachievement in gifted students is a topic of concern for educators, the onset of underachievement has been the focus of some researchers. Studies (Lupart & Pyryt, 1996; Peterson & Colangelo, 1996) have found grades seven through nine as a time of academic vulnerability, with seventh grade showing evidence of the greatest number of underachievers. Baker, Bridger, and Evans (1998) found that participants in their study began underachievement in late elementary school or in grades six, seven, or eight. They also determined that most of these students did not improve their academic performance during high school, and some even dropped out altogether. This is particularly relevant since Davis and Rimm (1998) note that 10 to 20 percent of high school dropouts test in the gifted range. There is a vast body of research literature concerning underachievement in gifted students. Although it is difficult to generalize across research with different methodologies and sample groups, many studies note that gifted underachievers have an external locus of control (Freeman, 1994), poor attitudes toward school, lack of motivation tn cooperate in the classroom (Seeley, 1993), and are less persistent than their achieving gifted peers (Gonzalez & Hayes, 1988). There are a variety of opinions concerning the severity of the underachievement problem. Studies place the percentage of gifted underachievers at different levels from as low as 15% (Seeley) to as high as 50% (National Commission on Excellence in Education, 1983). However, there does seem to be some agreement concerning the importance of learning style in the development of underachievement among gifted students.

Results of a study by Gohm, Humphreys, and Yao (1998) indicated that students gifted in spatial ability underachieve due to an incompatibility between the traditional curriculum and their individual learning style and interests. Their research revealed that students were less motivated by their educational experiences, performed below potential, and received less guidance from counselors concerning college preparation. Additionally, they noted that spatially gifted learners required more tactile activities coupled with less emphasis on lecture and drill and recitation.

Research by Whitmore (1986) on underachievement confirmed findings about the mismatch between student learning needs and school environment. She echoed conclusions by other researchers indicating how students turn to daydreaming, social interaction, activities outside of school, and even misbehavior to escape unrewarding classroom environments that are too easy, boring, and unsatisfying. She added that lack of motivation to excel is often the result of a mismatch between the type of activities provided in the classroom and the child's personal characteristics. Reis, Hebert, Diaz, Maxfield, and Ratley (as cited in Reis & McCoach, 2000) noted that boredom leads to lack of motivation when teachers expect students to complete assignments regarding concepts and content they mastered in previous years. Restak (1979) found that 60% of one's learning style is a biological and developmental set of characteristics that can contribute to effective instruction for some students, but ineffective instruction for those whose learning style does not match their school environment. Findings of several studies suggest that underachieving students make significant gains in classroom performance when their learning style preferences are accommodated (Andrews, 1990; Gadwa & Griggs, 1985; Klavas, 1993). Researchers add that students can gain from discovering their learning style. This knowledge allows students to fully utilize favorable conditions to enhance their learning, thereby taking more responsibility for their own progress (Dunn, Griggs, Olson, Beasley, & Gorman, 1995).

According to Dunn and Dunn (1993), learning style is the way individuals process, internalize, and retain new and difficult information. Therefore, focus on learning style means giving attention to individual strengths and differences during the learning process. Dunn, Dunn, and Price (2000) developed the Learning Style Inventory (LSI), a self-reporting instrument, to identify student learning style preferences. The LSI has been revised several times, most recently in 2000. By quantifying student responses to 104 statements on a five-point Likert scale this instrument provides educators with valuable information about the conditions under which each student learns effectively.

Purpose of the Study

The purpose of this study was to develop an instrument that would quantify student perceptions of the classroom environment. Along with the scores from the LSI, findings from a student perception instrument could answer the question, "How does the learning environment of the classroom, based on the students' perceptions of the classroom environment, match with the learning style preferences of the students?" Development of this instrument, the Student Perception Inventory (SPI), could reveal a student's perception of classroom elements that corresponds to her or his learning style. Use of such an instrument would make it possible to determine whether the classroom environment is compatible with the learning style preferences of any student, achieving or underachieving. For example, if a student's LSI indicates a strong preference for a quiet learning environment, the SPI could reveal how the student perceived the level of classroom noise. Comparing the student's scores from the LSI and SPI would indicate the degree of perceived compatibility with the learning environment. This information would be useful for teachers responding to individual differences in a child-centered classroom. With the ultimate goal of improving achievement and classroom performance, teachers would have valuable information to accommodate students' learning style needs.

Context of Instrument Development

Interest in the problem of under-achievement among gifted students has prompted this investigation into relationships between learning style and the classroom environment. Because gifted students have the potential to perform at high levels, they are expected to earn relatively high grades. According to Reis and McCoach (2000), when gifted students perform only at grade level, this should signal cause for concern.

Over the last decade, educational literature has focused on differentiating instruction to meet individual learner needs. Recognition that "one size" of instruction does not "fit" all learners has increased interest in identifying individual student learning styles so that student achievement can be maximized. Dunn, Dunn, and Price (2000) developed their learning style inventory based on several theoretical assumptions. They believe that individuals have learning style preferences that, when accommodated, can result in increased academic achievement and improved attitudes toward learning. Additionally, the stronger the preference, the more important it is to accommodate these learning styles. With more information about their students, teachers can make instruction compatible with a variety of learning styles through various classroom strategies, activities, and alterations of the classroom environment. Some researchers note that accommodating learning style preferences of students who are academically less successful is essential (Dunn, Griggs, Olson, Gorman, & Beasley, 1995).

The Dunn instrument (Dunn, Dunn, & Price, 2000) of learning style preferences (the LSI) has evolved over the years. The current version contains 104 statements with responses of "Strongly Disagree," "Disagree," "Uncertain," "Agree," and "Strongly Agree." Once a student completes the inventory, raw scores are converted to T-scores, which indicate the strength of the preference in each of 22 elements. The LSI indicates students" preferences in the areas of environment (noise levels, temperature levels, dim vs. bright lighting, and formal vs. informal classroom seating design), emotionality (presence of motivation, responsibility, and persistence; the need for structure or flexibility), sociological needs (learning alone or with peers, and/or in several ways, need for presence of authority figures, presence of parent motivation, and/or teacher motivation), and physical needs (auditory, visual, tactile, and/or kinesthetic learning modalities; learning in the evening or morning, late morning, or afternoon; need for intake of food or drink, and mobility in the classroom).

Scoring for the LSI, expressed as T-scores, ranges from 20 to 80 with a mean of 50 and a standard deviation of 10. A score from 20 to 40 indicates a low or opposite preference, while a score between 60 and 80 indicates a high preference. A score between 40 and 60 indicates that the element may not be critical to the individual's learning process, but will vary based on the student's interest in what is being learned and the difficulty level of the material (Price & Dunn, 1997).

A sample item from the LSI includes, "I study best when it is quiet." The response to this item, and others concerning noise, can reveal the need for a particular level of sound when dealing with new or difficult information. However, even when all of the LSI preferences are known, the educator does not know how the student judges or perceives those specific elements in the classroom. The level of noise in a specific classroom may be perceived as quiet by one student, but too noisy for concentration by another. Information provided by the LSI alone does not indicate compatibility with the classroom environment. Development of an instrument to measure student perceptions of the classroom environment would allow for this comparison.

Development of the SPI

This article describes development and field-testing of the SPI as a part of a larger research effort designed to examine the relationship between learning style preferences of gifted middle school students and their perceptions of the classroom environment and the impact this compatibility has on their classroom performance. This investigation has resulted in the development of a useful instrument to measure student perceptions of the classroom environment.

SPI content validity. Initial development of the Student Perception Inventory began with an examination of the LSI and the items it contains. Although LSI statements require students to indicate their preferences, the SPI was designed to obtain student perceptions of the classroom environment. Statements were worded to reveal the student's awareness of his or her environment regarding level of noise, lighting, temperature, classroom seating design (formal or informal), motivation, persistence level, structure in assignments, learning alone or with peers, presence of authority figures, auditory learning, visual learning, tactile learning, and kinesthetic learning activities, mobility in the classroom, and teacher motivation. A panel of education experts from the areas of curriculum, psychology and guidance, secondary education, research, and gifted education established the content-related validity for the SPI. These experts determined that the SPI should follow the same format as the LSI, with each statement followed by the same five Likert-type responses. Also like the LSI, the SPI requires students to respond to statements concerning an element from opposite perspectives. For example, with regard to temperature, the LSI requires students to respond to the statement, "I think best when I feel warm," while it later calls for a response to the opposite statement for temperature, "I usually feel more comfortable in cool weather than I do in warm weather."

Fifteen of the 22 LSI elements were used in the SPI. Seven LSI elements dealing with learning at different times of the day, eating and drinking while learning, and parent motivation were not included in the SPI. Because students do not have a choice about the time of day they are assigned to a particular class or the option of eating or drinking in the classroom, these areas were not considered in determining student perceptions. Reference to parents was also omitted as parents are not present in the classroom. The panel clarified wording of the SPI to ensure each statement called for student perceptions of the classroom environment rather than individual learning preferences. The SPI was pilot tested to further assess validity and to determine reliability.

SPI pilot testing and reliability. A reliability analysis was conducted after administering a pilot test of the SPI to 45 middle school students. This administration of the instrument included sixth, seventh, and eighth grade participants from three advanced homerooms made up of nongifted and gifted students. Language arts was the content area for all three grade groups. The 30-item instrument was administered to the students so that internal reliability could be established. Because there are two comparable statements for each of the 15 LSI elements on the SPI, a split-half reliability was determined. Raw scores were entered into the Statistical Program for the Social Sciences (SPSS, 1998) and a correlation performed on each element pair. The Pearson correlation coefficients ranged from .662 to .868, establishing internal consistency reliability. The Spearman-Brown correction was calculated for each of the 15 correlation coefficients and produced coefficients ranging from .796 to .929.

SPI scoring. The SPI contains two statements related to each of the 15 LSI preference elements, resulting in 30 statements. As in the LSI, students have the choice of selecting a response on the SPI for each statement of "Strongly Disagree," "Disagree," "Uncertain," "Agree," or "Strongly Agree." A raw score of 1 to 5 was assigned to each response. Combining the raw scores on the two SPI statements produced a possible raw score range of 2 to 10 for each SPI element. The SPI scoring followed the LSI convention of assigning scores for the responses based on the content of the statement. For stone statements, "Strongly Disagree" was assigned a value of "1" and "Strongly Agree" was assigned a value of "5," while on other statements "Strongly Disagree" was assigned a value of "5" and "'Strongly Agree" was assigned a value of "1." The determination for this forward or reverse scoring on the SPI was based on the degree to which the statement indicated the student's perception of the environment and in order to be compatible with the LSI scoring. For example, if the student responded "Agree" to the statement on the SPI, "When I am in my math class, I usually stay in my seat," the response earned a score of "2." If the student responded "Disagree" to the reverse statement, "When I am in my math class, I often move around the classroom," the response also earned a score of "2." The student's combined score for these two mobility statements was "4," out of a possible range of 2 to 10. A score of "4" indicated that, when involved in that specific class, the student did not perceive that he or she moved about the classroom. A student with little or no preference for mobility scored between 20 and 40 on the LSI. A low score on the SPI indicated the student perceived little or no actual movement in the classroom. Low scores for the mobility element on both the LSI and SPI indicated that learning style preference and the student's actual experience in the classroom were compatible. On the other hand, a student with a high need for mobility, as indicated by a score of 60 or above on the LSI, who scored a "4" for mobility on the SPI, revealed that the classroom was incompatible with his or her preference for movement.

As indicated above, the two statements for each element were written in an opposite manner. For example, the positive statement for persistence was, "When I am in my science class, l usually finish my class work without being reminded." The negative statement for persistence was, "When I am in my science class, I have to be reminded to continue with my class work." If a student responded "Agree" to both of these statements, the positive statement was scored a "4" and the negative statement raw score was "2." Agreement to opposite statements indicated the student was unsure of his or her perception or inadvertently marked the response incorrectly. When students' raw scores were two or more points apart for the same element, raw scores were removed from further calculations. This procedure resulted in removal of only 98 raw scores out of 6,000 paired responses.

The LSI generates both raw scores and T-scores; therefore, the SPI raw scores were converted to z-scores and then to T-scores. Correlating the LSI T-scores and SPI T-scores for each element permitted the examination of the degree of compatibility between students' learning style preferences and their perception of the learning style elements within the classroom. A compatibility index for each of the 15 elements was computed by subtracting the SPI T-scores for each content area from the corresponding LSI T-scores. The compatibility index was then correlated with the students' specific content area grades indicating whether the degree of compatibility between preference and perception had an impact on students' performance.

Method

Participants

From a total of 90 identified gifted sixth, seventh, and eighth grade students, 80 gifted students from two middle schools participated in the field test after parental approval was obtained. These gifted students were from one school district in a Southwest Georgia community. The gifted participants included 26 sixth graders, 34 seventh graders, and 20 eighth graders. All of the participants were identified as gifted during elementary school by meeting established criteria in the following categories: mental ability (at or above the 96th percentile on The Otis-Lennon School Ability Test), achievement (at or above the 90th percentile on either the Iowa Test of Basic Skills or the Stanford Achievement Test [9th ed.]), creativity (at or above the 90th percentile on the Torrance Test of Creative Thinking), and motivation (an overall grade point average of 3.5 on a 4.0 scale). The classroom teachers of these 80 gifted students also participated in the study because the SPI was administered in their classrooms. Fifty-six teachers in all were involved in administration of the SPI. Six of these teachers had training for teaching gifted students. Teaching experience ranged from 1 to 24 years with the average being 12.6 years.

The gifted students who were identified as underachieving had an overall numerical average of 84 or below for their academic subjects in reading, language arts, social studies, math and science. Sixteen of the 80 students (20%) fell within this grade range. Underachievement among all gifted students is estimated by Seeley (1993) to be from 15 to 40%. Of the 16 underachieving students, 2 were sixth graders, 9 were seventh graders, and 5 were eighth graders.

Procedure

The SPI was field tested in April of the 2000-2001 school year, following administration of the LSI in March. Students completed one SPI for each of their academic classrooms, which included reading, language arts, math, science, and social studies. It was necessary to obtain perceptions from students concerning the classroom environments for each content area because student performance levels often varied from content area to content area.

In early April, the researcher administered the first of five SPI instruments. This involved nine classrooms in the two participating schools. In the initial administration, the purpose of the SPI was discussed with the students. Because gifted students are keen observers and have intuitive abilities (Davis & Rimm, 1998), it was decided to bring them into the process as partners in research. All student participants had received training in the scientific method of investigation, engaged in science fair projects, and had knowledge of the research process.

Classroom teachers conducted subsequent administrations of the SPI during the time when students were assigned to their classrooms. Because the students were given extensive directions by the researcher during the initial administration of the SPI, the teachers only needed to issue a reminder.

Students required 12 to 15 minutes to complete each SPI. After SPI administration, each classroom teacher returned the instruments to the school office for collection by the researcher. A raw score was assigned to each response for each of the 30 statements. Scoring of each SPI required about 1 to 2 minutes and was completed by the researcher.

Results

Table 1 displays the LSI means for the achievers and the underachievers. Examination of this table reveals that underachievers had a strong preference for dim lighting in their environment as indicated by a mean score of 39.88. Dunn, Dunn, and Price (2000) indicate that T-scores below 40 show a strong preference for dim lighting, while scores above 60 show a strong preference for bright lighting. Lighting needs for achievers, as indicated by a mean of 45.89, revealed that these students did not have a strong preference in the lighting category, but tended toward less bright lighting. Another area that revealed a strong preference by the underachievers was in the area of tactile learning. Their mean score of 62.81 revealed a strong preference for hands-on learning activities since scores above 60 indicate a strong preference for this modality. Achievers had a tactile mean of 56.77. While this does not show a strong preference, scores above the mean of 50 tend toward a preference for this type of learning.

By comparing the results of the LS1 and the SPI for low achieving students, this study revealed which classroom elements needed to be altered for student success. For example, tactile learning was highly preferred by low achievers but they perceived that few tactile activities were used in their classrooms. Likewise, these students also prefenced kinesthetic modalities, but perceived its wide use only in math classrooms.

Most of the gifted students managed to persist and perform at high levels even though their environments were not always compatible. However, it is possible the low achieving gifted students were not able to overcome incompatibility with the classroom and performed below their potential. See Table 2 for a summary of the classrooms revealing a significant degree of compatibility between learning style elements and the students' perception of the element within the classroom. The SPI indicated that participants perceived only a few elements within their classrooms as compatible with their learning style preferences. Out of 15 LSI elements, only 7 preferences were perceived to be compatible in one or more of the classrooms. Math was the only content area compatible in 6 of these 7 learning style preferences.

SPI validity and reliability. A testretest was conducted with 20 of the participants by comparing the T-scores of two administrations of the SPI. Students' initial responses to the SPI statements in their science class were correlated with an additional administration of the SPI in the same science class 4 weeks later. Comparisons of these two sets of T-scores produced correlation coefficients for each of the 15 SPI elements. The range of coefficients was from .808 to .977. The range of coefficients, when corrected by the Spearman-Brown formula, was from .893 to .988.

Discussion and Implications

The LSI is a valuable tool educators can use to determine student learning style preferences. However, knowing student preferences only provides teachers with partial knowledge of their students' learning needs. Students not only have different learning styles, but they also have different perceptions of conditions within the same classroom. What may be a warm room for one child may be too cold to concentrate for another. Or within the same classroom, one student may perceive that he or she has adequate opportunity for hands-on activities, while another child may need and want more tactile experiences. Conditions may also change from one classroom to another as a result of teacher preferences or other environmental circumstances. Currently, there is no instrument that asks students to judge the environment of a specific classroom in order to determine the degree of compatibility with their learning style preferences. Development of this perception instrument is a step in the right direction.

Information provided by the SPI in this study was informative and useful. Teachers in any content area can use the SPI with any group of students. It is easy to administer and score. Using the SPI with the LSI, teachers can determine which elements in their classroom are compatible with student learning styles and which elements need to be adjusted for more effective learning. This type of information would allow a teacher to determine specific accommodations necessary for a particular student.

While the SPI may be useful to any classroom teacher, it was primarily developed as a method to determine compatibility between gifted students' learning style preferences and their perception of actual classroom environments. However, it could also be used in research to determine differences in student perceptions by gender, age, ethnicity, or achievement levels. This instrument is flexible enough to be used with an individual student, several classroom groups, or an entire school population. The SPI could also be modified by adding more questions to compare other perception elements that affect student learning.

Although underachievement is a complex problem, teachers can gain valuable information about low achievers if they utilize instruments like the LSI and the SPI. By knowing how low-achieving students learn and under what conditions, educators will be better equipped to alter the classroom in order to meet individual learning needs. Teachers must take advantage of any opportunity to reach their students more effectively, especially underachieving students. If students are not learning the way teachers are currently teaching, then they should make every possible effort to teach the way students learn. In other words, educators should provide environments and use strategies that allow all students, gifted and nongifted, to learn in the most effective manner. Although gifted students are not usually viewed as needing accommodations, they need and deserve to have effective learning environments as much as any other learner.

Table 1
LSI Descriptive Statistics Summary

                    Range for All Learners   Underachievers
                           Achievers

LSI Elements         Min.    Max.    Mean         Mean
                                     n=64         n=16

Sound Level         31.00   74.00   49.98        52.00
Lighting             9.00   66.00   45.89        39.88
Temperature         20.00   71.00   49.45        50.19
Classroom
Seating Design      33.00   73.00   44.05        45.81
Motivation          20.00   69.00   49.61        51.88
Persistence         20.00   65.00   49.73        44.75
Structure in
Assignments         31.00   73.00   50.69        54.25
Alone/with Peers    27.00   77.00   51.78        48.94
Authority Figures
Present             25.00   74.00   47.56        48.13
Auditory            20.00   69.00   47.80        49.69
Visual              26.00   71.00   49.69        52.75
Tactile             34.00   69.00   56.77        62.81
Kinesthetic         26.00   75.00   55.69        59.5
Mobility            22.00   65.00   53.55        48.63
Teacher Motivated   25.00   68.00   51.50        51.94

Table 2
Degree of Compatibility between
Learning Style and Perception

LSI and SPI Elements /                           Pearson Correlation
Content Area Classroom                                Coefficient

Persistence / Reading                                   .510 **
Persistence/ Language Arts                              .554 **
Persistence / Science                                   .544 **
Persistence / Math                                      .270 *
Persistence / Social Studies                            .390 **
Mobility / Language Arts                                .243 *
Mobility / Math                                         .397 **
Mobility / Social Studies                               .257 *
Self-Motivation / Reading                               .352 **
Self-Motivation / Language Arts                         .263 *
Self-Motivation / Math                                  .245 *
Teacher Motivation / Science                            .283 *
Teacher Motivation / Math                               .348 **
Working with Peers / Math                               .461 **
Presence of Authority Figures / Social Studies         -.279 *
Kinesthetic Learning / Math                             .247 *

* Correlation is significant at the 0.05 level (2-tailed)

** Correlation is significant at the 0.01 level (2-tailed)

REFERENCES

Andrews, R. H. (1990). The development of a learning styles program in a low socioeconomic, under-achieving North Carolina elementary school. Journal of Reading, Writing, and Learning Disabilities International, 6, 307-314,

Baker, J. A., Bridger, R., & Evans, K. (1998). Models of underachievement among gifted preadolescents: The role of personal, family, and school factors. Gifted Child Quarterly, 42, 5-14.

Baum, S. M., Renzulli, J. S., & Hebert, T. P. (1995). Reversing underachievement: Creative productivity as a systematic intervention. Gifted Child Quarterly, 39, 224-235.

Davis, G. A., & Rimm, S. B. (1998). Education of the gifted and talented (4th ed.). Needham Heights, MA: Allyn and Bacon.

Dunn, R., & Dunn, K. (1993). Teaching secondary students through their individual learning styles: Practical approaches for grades 7-12. Boston: Allyn and Bacon.

Dunn, R., Dunn, K., & Price, G. (2000). Learning Style Inventory. Lawrence, KS: Price Systems.

Dunn, R, Griggs, S. A., Olson, J., Gorman, B., & Beasley, M. (1995). A meta-analytic validation of the Dunn and Dunn model of learning-style preferences. Journal of Educational Research, 88, 353-361.

Emerick, L. J. (1992). Academic underachievement among the gifted: Students' perceptions of factors that reverse the pattern. Gifted Child Quarterly, 36, 140-146.

Freeman, J. (1994). Some emotional aspects of being gifted. Journal for the Education of the Gifted, 17, 180-197.

Gadwa, K., & Griggs. S. A. (1985). The school dropout: Implications for counselors. The School Counselor, 33, 9-17.

Gohm, E. C., Humphreys, L. G., & Yao, G. (1998). Underachievement among spatially gifted students. American Educational Research Journal, 35, 515-531.

Gonzalez, J., & Hayes, A. (1988). Psychosocial aspects of the development of gifted underachievers: Review and implications. The Exceptional Child, 35, 39-51.

Gowan, J. C. (1955). The underachieving gifted child: A problem for everyone. The Exceptional Children, 21, 247-249, 270-271.

Klavas, A. (1993). In Greensboro, North Carolina: Learning style program boosts achievement and test scores. The Clearing House, 67, 149-151.

Lupart, J. L. & Pyryt, M. S. (1996). "Hidden gifted" students: Underachiever prevalence and profile. Journal for the Education of the Gifted, 20, 36-52.

National Commission on Excellence in Education (1983). A nation at risk: The imperative for educational reform. Washington: U. S. Department of Education.

Peterson, J. S., & Colangelo, N. (1996). Gifted achievers and underachievers: A comparison of patterns found in school files. Journal of Counseling and Development, 74, 399407.

Price, G. E., & Dunn, R. (1997). Learning Style Inventory (LSI): An inventory for the identification of how individuals in grades 3 through 12 prefer to learn. Lawrence, KS: Price Systems.

Reis, S. M., & McCoach, D. B. (2000). The under-achievement of gifted students: What do we know and where do we go? Gifted Child Quarterly, 44, 152-170.

Restak, R. (1979). The brain: The last frontier. New York: Doubleday.

Seeley, K. R. (1993). Gifted students as risk. In L. K. Silverman (Ed.), Counseling the gifted and talented (pp. 263-276). Denver: Love Publishing.

SPSS Graduate Pack 9.0 for Windows [Computer Software]. (1998). Chicago, IL: SPSS.

Starnes, W. T. (1988, April). A study in the identification, differential diagnosis, and remediation of underachieving highly able students. Paper presentation. Annual convention of the Council for Exceptional Children. Washington, DC.

Supplee. P. L. (1990). Reaching the gifted under-achiever. New York: Teachers College Press.

Whitmore, J. R. (1980). Giftedness, conflict, and underachievement. Boston: Allyn and Bacon.

Whitmore, J. R. (1986). Understanding a lack of motivation to excel. Gifted Child Quarterly, 30, 66-69.

Letty J. Rayneri is the director of gifted education for the Dougherty County School System in Albany, Georgia. She is responsible for coordinating the curriculum for the district's gifted students in grades 1-12, as well as providing teacher training on the needs of gifted students. She completed her Doctor of Education degree at Valdosta State University, where she focused her dissertation on the learning styles and needs of gifted middle school students. Dr. Rayneri is president of Georgia Association for Gifted Children. E-mail: [e-mail]

Brian L. Gerber is an associate professor of science education and associate dean of the College of Education at Valdosta State University in Valdosta, Georgia. His research interests include matching instruction to the needs of students, informal science learning, research-based teacher professional development, and technology integration in the classroom. E-mail: [e-mail]

Manuscript submitted February 26, 2003.

Revision accepted April 13, 2003.