A Study of Accelerated Reader
Model and Master Schools and
Performance on the Mississippi Curriculum Reference Test
A Quantitative Study
By Cindy Harper
For the
Masters Degree in Library and Information Science
Summer 2001
228-497-4509
Table of Contents
Research Problem page 3
Definitions page 3
Importance of the Study page 4
Hypothesis page 5
Limitations page 5
Related Literature page 6
Methodology page 8
Presentation of the Data page 9
Reading Scaled Scores page 9
Acceptable Reading Percentages page 11
Language Scaled Scores page 13
Mathematics Scaled Scores page 15
Scores within School Districts page 16
Conclusion page 20
Cited Sources page 22
Charts
Graphs
Research Problem
The purpose of this paper was to determine if the Mississippi schools that certified as either “Model Accelerated Reader School” or “Master Accelerated Reader School” (in the 2000 – 2001 school year) scored higher than the state average on the Mississippi Curriculum Test (MCT) that was administered in spring 2001.
Definition of Accelerated Reader
Accelerated Reader (AR) is an Advantage Learning Company computer program, providing quizzes that are used to test students on reading comprehension of books. The program ranks books for difficulty of reading on an ATOS formula taking into account the length and difficulty of words, sentences and paragraphs. Students are tested and placed on individual reading levels through administration of a STAR, test also marketed by Advantage Learning. Students select and read books from school and classroom libraries. Points are assigned to the books based on the amount of time a student is expected to need to finish the book. Thus more difficult and thicker books carry higher point values. Simple picture books are assigned a point value of .05; transitional or simple chapter books on a high second to third grade level are assigned a one point value; more complex chapter books intended for readers above third grade level vary in point value. As an example, the Goosebump series has three points per book while Harry Potter and the Sorcerer’s Stone is worth twelve points. Students are awarded full point value only if they answer each question correctly. No points are awarded for scores of 60% or lower.
Importance of the Study
Mississippi schools using Accelerated Reader have a substantial financial investment in the program. Schools using Accelerated Reader must already have computers in place. There should be at least one computer to a classroom to allow students to take an AR quiz after completing a book. The initial investment can range from $499 for a starter set for 250 students that includes the program software plus four reading title sets containing up to 200 quizzes to $1499 for a set with 20 title disks or 1000 quizzes. A super kit that contains the STAR software and costs $2999 is also available. Once a school has established the Accelerated Reader program, it will need to continue to add disks that cost $87 each for approximately 20 – 50 quizzes or purchase quizzes individually at the rate of $2.58 each. School and classroom libraries must also provide sufficient numbers of books per student on each reading level. Schools may also purchase additional components including an evaluation program for early literacy of pre-kindergarten through second grade students and a program that tests students on literacy skills.
Advantage Learning provides seminars and support techniques entitled Reading Renaissance to help teachers take full advantage of the program’s components. Advantage Learning’s literature suggests that the use of the Accelerated Reader Program will cause standardized test scores to go up in not only reading but other areas as well. The company has in place a certification process to recognize chools that employ a majority of the Reading Renaissance techniques. Schools that complete a survey and show sufficient student points (equating to amount of time given to reading practice) and percentages of 85% or better (equating to quality of reading comprehension) with no more than 15% of students at risk (equating to expected performance) for a grading period are certified as AR Model schools. Maintaining the same standards for a continuous 18-week period plus showing reading growth may result in AR Master Certification. All grade levels at the school must be included. A school must seek recertification in succeeding years.
Assumptions of the Study
For this study, it was assumed that schools certified as either Model or Master used the Reading Renaissance teaching methods
Hypothesis
. The hypothesis of this study was that AR Model and Master Schools would score higher than the Mississippi mean score on the MCT’s reading scaled score. A secondary hypothesis was that language and mathematics scaled scores would also be higher for AR Model or Master schools, as reading comprehension would help students with comprehension in all subject areas.
Limitations
The MCT was chosen as the comparison document, as it was the only standardized testing instrument administered state-wide in the 2000 – 2001 school year. The MCT was administered for the first time in the 2000-2001 school year to second through eighth graders; hence the study was limited to that school year for students in those grades. As no Mississippi schools with students above sixth grade level received AR Model or Master certification, seventh and eight grades were not included in the study.
Related Literature
Much of the published research about the Accelerated Reader Program is funded or supported by Advantage Learning Company. Some independent journal articles include the information that standardized test scores did rise after implementation of Reading Renaissance techniques. The rise in standardized test scores was the most relevant to this study, which examined scores from the Mississippi Curriculum Test and compared the results between Accelerated Reader Model schools and the Mississippi averages.
Fifth grade students in Tift County, Georgia scored 14 points higher on statewide reading tests after one year of using the Accelerated Reader Program (Lopez, 2000). The program had been implemented by Librarian Terri Nalls, who wanted to help students in Spencer Elementary School where 92% of the student population was low income and came from homes of 30 to 40% illiteracy. Another journal article was published the following year stating that the Tift County Schools had implemented two other reading programs: Reading Recovery (for children who need extra help) and Literacy Collaborative (long-term staff professional development). (Gomez, 2001)
Texas educators implemented Reading Renaissance techniques (including day long teacher training) in schools where half the students were bilingual, speaking only limited English. After using Accelerated Reader and Reading Renaissance, a larger number of students passed the Texas Assessment of Academic Skills (TAAS) in both reading and mathematics than they had before AR was used. In 1997, 76% of sixth graders passed the TAAS reading test as opposed to 1995 when only 50% of fourth graders passed the TAAS reading test. In TAAS mathematics, 47 % of the fourth graders passed in 1995 improving to a 91% passing rate as sixth graders in 1997. (Texas, 1998)
Deborah Biggers wrote a critical editorial in the Journal of Adolescent & Adult Literacy against the Accelerated Reader program. She was especially critical of 64 “scientific research” documents posted on Advantage Learning’s website. Of the research documents, 29 were school or district evaluations and16 were done by Advantage Learning itself, leaving only 19 independent studies. However, within the 19 independent studies, there were five studies authored or coauthored by K. J. Topping, who had also coauthored with T.D. Paul, the husband of founder Judy Paul (Biggers, 2001).
In one of the studies authored by Topping and Paul, 39 participating states were ranked in terms of student performance on the National Assessment of Educational Progress (NAEP) Study by the U.S. Department of Education. (Topping & Paul, 1999) In the 1994 NAEP study, Mississippi students were among the bottom quartile with an average 202 mean score as opposed to students in the top quartile with an average 221.4 mean score. Those in the middle quartile had an average 213 mean score. Topping and Paul correlated that the amount of time students spent reading was a factor in scoring higher on the NAEP and recorded the mean AR points for each of the groups. AR points are awarded based on the time spent reading; thus larger numbers of points equates to more time spent reading. Mean AR points were 39.2 in the top quartile, 25.1 in the middle quartile, and 24.7 in the low quartile.
Methodology
This study was a quantitative study that examined and compared two sets of documents. The Mississippi schools with AR certification were identified from a spread sheet obtained from Advantage Learning. MCT scores were obtained http://www.mde.k12.ms.us/account/rc02/HTML/RC02DIS3.HTM on-line at the Mississippi Department of Education’s web site.
Using the Reading Renaissance document, the researcher identified the 20 to 25 schools (dependent upon grade levels included in each school) that were to be included in the study. School districts vary in the grade levels taught in their individual schools. Some districts use lower and upper elementary schools. splitting K-2 and 3 – 5. Other school districts have schools that house just third and fourth graders or fifth graders only. Also, some school districts do not include sixth graders in elementary schools. The researcher included only the grades as identified by Reading Renaissance.
The MCT scores of the students in each individual grade were recorded on a spread sheet. The researcher included the name of the distinct, school and grade level. The data that was included for each grade level was the reading scaled score, and percentages of students who were ranked as Minimal, Basic, Proficient and Advanced in reading. Because Proficient and Advanced are considered acceptable levels of reading competency, the sum of the percentages was calculated for comparison purposes. Language and Mathematics scaled scores were recorded, but the percentages in the four categories of student performance were not, as this study emphasized reading scores. In addition, the number of schools (with the same grade level) within a district was counted using the State Department of Education’s MCT web pages.
Because the MCT data was arranged by grade level, the grade levels were compared individually in the study. The data was ranked first by reading scaled score, then by Acceptable percentages (the sum of Proficient and Advanced). Rankings of Language and Mathematics scaled scores were also made. The researcher also examined the reading scale performance of AR certified schools within their own school districts.
Presentation of the Data
Reading Scaled Scores
A high majority of the AR certified schools did score above the Mississippi mean on the reading scaled scores. More second graders in the study were above the Mississippi mean than were students in the other grades. The number of schools in the study scoring above the Mississippi mean was better for grades two and three than for four, five and six. In comparison, there was a larger difference in points scored for those schools above the mean than below the mean.
Ninety – two percent (23 of 25 schools) of second grade schools recorded a score higher than the Mississippi mean of 454. (Chart 1, Graph 1) The two schools that were below the Mississippi mean scored 5.2 and 26.9 points below the mean. The highest reading scale score was 490.1 or 36.1 points above the mean.
For the third grade, 84% (21of 25) of the studied schools were above the Mississippi mean of 478.5 (Chart 2, Graph 2) Manatachie school was only 2.5 points below the Mississippi mean. The lowest ranked school was 24.9 points lower than the Mississippi mean. Monroe third graders (the lowest ranked of those above the mean) scored 7 points higher than the mean. A Biloxi school scored the highest score of 514.8 or 36.3 points higher than the mean.
Eighteen of 23 or 78% of the studied schools with fourth graders were above the mean of 500.4. (Chart 3, Graph 3) Of those below, two schools were only .1 and .2 points below the mean. The third school was close to the mean at 499.2 points or .9 points lower. The lowest ranked school was 11.1 points below the mean with a score of 489.3. Of those schools that scored better than the mean, East Jones was the closest to the mean but still scored three points above the mean. The highest ranked school was 33.4 points above the mean.
Eighty percent of the AR schools (16 of 20) with fifth graders were above the mean of 516.2. (Chart 4, Graph 4) Hopewell School of the Covington District was 25.6 points below the mean. Mason School of the Laurel District was also below the mean but was closer to it, being only 5.2 points lower. On the other end of the scale, Jeff Davis of the Biloxi District was 33.8 points higher than the mean.
The performance of sixth graders was better than that of fifth graders with 87.5% (14 of 16) of studied schools scoring above the Mississippi mean of 525.8 (Chart 5, Graph 5). Hunt International of the Columbus District was only .4 points below the mean, while Hopewell of the Covington District was much further below the average at 36.8 points lower. In the other direction, Monroe of the Hamilton District was closest to the mean and scored 5.3 points higher. Another Biloxi District School, Beauvoir, performed the best and scored 561.7 or 35.9 points higher.
Hopewell School of the Covington District consistently ranked lowest for all grades (two through six) in the study. At its best, Hopewell was 11.1 points below the mean, and at its worst, was 36.8 points below the mean. Hopewell had recertified as an AR Model School in the 2000 – 2001 school year. Mason School (grades K – 5) of the Laurel District also consistently ranked below the Mississippi mean in reading. Mason had certified as Model School that school year. Hunt Intermediate (grades 5 and 6) of the Columbus District also scored below the mean for both grades included in the study. Beauvoir School of the Biloxi District was among the top three highest scorers in all grades (two through six.) The performance of Beauviour students ranged from 24.9 points to 36.1 points greater than the Mississippi mean.
Acceptable Reading
Percentages
The State Department of Mississippi ranked the students by percentages in four quartiles. Minimal and Basic categories are considered to be unacceptable; and students ranked in those categories are not reading on grade level. Proficient is considered to show that the students are performing reading at grade level, while Advanced students are performing above grade level. For this study, the numbers in Proficient and Advanced were combined into a category labeled as “Acceptable”.
For the most part, the schools in the study did perform better than the Mississippi average. Those schools which had lower scaled scores often had lower percentages of students especially in the Advanced category but sometimes also in the Proficient category. There were some exceptions. The performance of sixth grade students at Hopewell school was especially low and alarming.
As with the Reading scaled score, 23 of the 25 second grades in the study were above the Mississippi average. (Graph 6) While the Mississippi average for Acceptable reading level was 78.8 %, the studied schools above the average ranged from 81% to 100%. The performance of Mason and Hopewell schools which were also lowest in Reading Scaled Score were 64.4% and 66.7% respectively.
Of the studied schools with third graders, 20 of 25 were better than the Mississippi average of 74.1% when Proficient and Advanced were combined. (Graph 7) Those above the state average ranged from 80% to 98.3%. Of the five below the state average, only one was more than 4% lower than the Mississippi average. Four of the five below the state average were close to the state average and to one another, ranging from 70.5% to 73.9%. Hopewell school was the lowest with only 58.8% students reading at Acceptable levels. Hopewell’s Minimal and Basic percentages were nearly the same 21.6% and 19.6%. At that school 54.9% were Proficient and only 3.9% were Advanced.
The performance of fourth graders in the Acceptable percentages was better in that category than it had been for Reading Scaled Score (18 of 23). Twenty-one of the studied schools were above the Mississippi 81.1% average (Graph 8). The percentage of fourth grade students in the AR Model and Master schools that scored above the state average ranged from 83.5% to a perfect 100% score. The two schools that were below the average were not significantly lower. Their averages were 75% and 80%.
Like the scaled scores, 16 of 20 fifth grade schools showed percentages better than the Mississippi average of 77% (Graph 9). For those schools scoring better than the state average the range was 80.4% to 96.7%. The top school, Beauviour, had only 1.7% in each Minimal and Basic category; 50% in Proficient, and 46.7% in Advanced. Of the schools below the average, Covington scored only 55.9% as Acceptable; the school’s Proficient category was the same as Beauvior with 50%. Covington’s Advanced category, however, had only 5.9%. Covington had a large percentage (29.4%) in the lowest category of Minimal. The other three schools below the Mississippi average were at 61.1%, 73.8%, and 75.4%.
Sixth graders in Mississippi showed a lower percentage of readers in the combined Proficient and Advanced categories than in lower grades. The Mississippi combined average for sixth graders was only 68%. Of the 16 schools in the study, 14 scored above the State average with scores ranging from 73.5% to 95.2%. (Graph 10) Hunt International at 66% was barely below the state average. Hopewell, on the other hand, was very low with only 32.4% considered as Acceptable; in fact, Hopewell had no students in the Advanced category. The school’s largest percentage (43.2%) was in the lowest Minimal category.
Language Scaled Scores
The language scaled scores did not match up to reading scaled scores. The percentage of AR certified schools above Mississippi’s mean language scaled scores was less than those for reading scaled scores. While the majority of schools in the study did score above the Mississippi mean for language scores, the percentage was not high as those recorded for reading scores.
Seventy-two percent (18 of 25) of second grade AR Model and Master schools were above the state language scaled score of 450.4. (Graph 11) Five schools that had scored above the Mississippi mean for reading did not do so for language. The lowest language score was 20.5 points below the mean; however, the highest language score was 41.2 points higher.
Twenty of 25 or 80% of third grade schools scored above the Mississippi mean of 479.9 in language scaled score. (Graph 12) While four of the schools also scored low in reading, an additional school, Singing River of the Pascagoula School District, was below the language mean. Singing River, however, was only .7 points lower than the Mississippi mean. The lowest score from Hopewell was 20.9 points lower than the mean. The school with the highest language scaled score was 42.8 points above the Mississippi mean.
In the study, 69.5% or 16 of 23 fourth grade schools scored above the Mississippi mean of 500.3. (Graph 13) In the reading scaled score, 78% were higher than the mean. The highest ranked school was 43.2 points above the mean. The range of scores below the mean, however, was not significantly lower. The lowest ranked school was only 6.1 points below the mean.
For the language scaled score, 70% of the fifth grade AR Model or Master schools were above the Mississippi mean of 518.9. (Graph 14) As in other grades, the studied schools did not do as well in language as they did in reading. The studied schools had 80% above the state’s mean for reading scores. The highest scoring school for language was 35.8 points higher, while the lowest scoring school was 25.1 points below the mean.
The sixth graders performed the best in language scaled scores of all the grades in the study. Ninety-four percent or 15 of 16 schools were above the Mississippi mean of 525 for language. (Graph 15) The highest language score in the study was 31.1 points higher than the Mississippi mean. The only score lower than the mean, however, was considerably lower by 38.4 points. Hunt International of Columbus was above the Mississippi average for the first time in the study, having scored .4 points better than the state language mean.
Mathematics Scaled
Scores
The AR Model and Master schools did not do as well in mathematics as they had in either reading or language. Some of the same schools were below the state mean in both mathematics and language.
Seventy-six percent (19 of 25) of second grade schools in the study were above the Mississippi mean of 428.4 in mathematics. (Graph 16) The highest scoring school was a full fifty points above the state mean. The lowest ranked school was 38.2 points below the mean.
For third grade schools, the percentage was also 76% (19 of 25) for mathematics scaled scores above the state mean. (Graph 17) The highest ranked school scored 42.8 points above the state mean of 474.5. The lowest ranked school was 29.3 points below the mean.
Only 74% of the study’s fourth grade schools were above the state’s 500.1 mathematics scaled score. (Graph 18) The top ranked school was 43.2 points higher while the lowest ranked was 17.2 points lower.
In the fifth grade schools, 75% (15 of 20) of the AR certified schools scored higher than 520 which was the Mississippi mean in mathematics. (Graph 19) The highest mathematics score was 45.6 points above the mean. The lowest score was only 8.8 points below the mean.
The mathematics performance of sixth graders in the study was consistent with their reading scores. Fourteen of sixteen, or 87.5% of the sixth grade schools were above the state’s mathematics mean of 540 (Graph 20) The sixth graders were also 87.5% above in reading, but were 94% above in language. In mathematics, Beauviour students scored 45.7 points higher than the Mississippi mean; while Hopewell students scored 34.2 points below. Hunt International, which was also below the mean, was only 3.5 points lower.
Scores within
Schools with Good
Performances
Biloxi school system was one of two to have all its elementary schools certify as either AR Model or Master. Beauvior and North Bay were consistently among the top performers in all subject areas of the MCT. The district’s reading scaled scores were well above the Mississippi mean, but there were some differences among the schools. The difference in reading scores among the six Biloxi schools was as much as thirty points. Monroe School District had all four of its schools with grades two through six to certify. The Monroe District and its individual schools were above the state average in all areas studied. (Chart 21)
In the Amory School District, only West Amory certified with second graders. With only two schools housing second graders in Amory, West Amory School was equal to the district in reading scaled score and above the state mean by 6.7 points. Gulfport District’s Bayou View School with grades two through five was an AR Model School. Gulfport with seven elementary schools scored above the state mean in reading for all grade levels. Bayou View School students scored above the district’s average scores in all subjects in all grades. Caledonia School with grades 2, 3 and 4 was the only one of three Lowndes District schools to certify as an AR model school. All grades of the Caledonia School were above the district average and the district was above the state reading mean. Taconi Elementary School which houses all the Ocean Springs School District’s fifth graders was 29.7 points above the Mississippi reading mean. Lower elementary schools in Ocean Springs did not certify and were not included in the study.
Schools with Mixed
Results
Jones District had five or six (depending upon grade level) elementary schools, but only East Jones Elementary school was an AR Model school. Jones District and East Jones Elementary School each scored above the state mean in all subject scores but language in the fifth and sixth grades. Although East Jones was an AR Model school, its individual scores were below the district’s averages in all subjects (including reading) for most grades. The Pascagoula School District has eleven elementary school housing grades kindergarten to five. Only Singing River was certified. Singing River’s fourth graders were 2.1 points lower than the district reading mean. Grades Two and Three were about 3 to 4 points higher than the district mean, while fifth graders were 7.9 points higher. The reading scale score of the district was higher than the Mississippi average. Rankin District was above the state mean. In its six schools with second grades, three were AR certified, but only one (Flowood) was above the district average. For third graders, 4 of 6 schools were certified. Richland and Flowood were below the district average. Rankin had two certified schools with fourth graders among its four (4th grade) schools; both were above the district average. In the Tupelo District with six schools for grades 2 – 4, only Thomas Street was studied as a recertified AR Model School. The Thomas Street second and third graders were slightly below (2 to 4 points) the district reading average. The school’s fourth graders were .4 points above the district average. Thomas Street and Tupelo District were above the state mean.
Mason Elementary School held second through fifth graders in the Laurel School District that had four schools for comparison purposes. (chart below) Nora Davis Magnet School in the same school district had second through sixth graders. The district had only one other school housing sixth graders (it was not an AR certified school). The school district scored above the state in only reading for third graders. Nora Davis Magnet school, however, scored above the state average (and well above the district’s average) in all areas. Mason’s scores were below the state average, but above the district average.
Schools with Poor
Performances
Manatachie School with grades two through six in the Itawamba School district performed poorly. Manatachie School’s average was slightly below the state’s in all reading scores with the exception of second grade where the students were only .7 points above. Manatachie’s fourth grade scored .3 below the state’s reading mean, the third graders scored 2.6 below and the fifth graders scored 2.7 below. The greatest difference was in sixth grade where the students were 11.9 points below the state reading mean. While Manatachie scored below the state, the Itawamba District did not. There were four second through fifth grade and three sixth grade schools in the district. Manatachie is a Master School which indicates it had shown improvement in reading scores during the school year. The language and math scores for Manatachie and the Itawamba School District followed about the same pattern as reading scores, with Manatachie slightly below the state mean, and the district above the mean. (Chart 21)
In the Covington School District, Hopewell School scored below the state average in all areas. Hopewell, a recertified AR Model School, was also below the district’s averages for all grades. (Graph below) Two other schools in Covington were in the study. Collins was a recertified Master School (indicating reading growth within the school year), yet it scored 6.3 points below the state’s reading mean for third graders and 3.8 below for fourth graders. Seminary scored only .1 points below the state mean with fifth graders but above the in all other grades was a first time AR Model School. A fourth Covington District School had not received AR certification and was not studied.
Huntwell Elementary School in the Columbus District had also scored consistently below the state average, but there were no other schools with 5th and 6th grades in that district; so comparisons could not be made.
Conclusion
The majority of schools that had certified as AR Model and Master Schools did score above the Mississippi mean on the MCT reading scaled score given in the spring of 2001. The study also showed that all AR certified schools had more than 50% of students reading at or above grade level. The percentages of students reading at Proficient or Advanced for the majority of AR certified schools were higher than the Mississippi average.
Most of the AR certified schools also scored above the Mississippi mean in language. The language scores of the AR Model and Master certified Schools, however, were not as high as the reading scores. The majority of the studied schools also scored above the state’s mathematics scaled score mean. There was a larger number of studied schools scoring below the math mean than in either language or reading. The results suggest that Reading Renaissance is not as effective in raising language and math scores as it is in elevating reading scores.
Overall, second graders did better than students in upper grades. This may suggest that intervention and enhancement in early grades is more effective in raising reading comprehension scores than in latter grades . A follow-up study should be done to see if upper grades will improve as the schools continue to use Accelerated Reader.
There were a few disturbing results that need further study. Because Hopewell School of the Covington School District consistently scored below the state and its district, it should be looked at for other factors. It may be that the district’s schools are not equally balanced in socioeconomic factors. The same may be true of Manatachie School in the Itawamba District. Itawamba was one of the few AR Master schools in the study. One of the criteria for Master certification was to show improvement in reading scores for the school year, yet Manatachie’s third and fourth grades were below the state’s reading mean. In the Covington School District, Collins was a recertified Master School, yet was below the district averages. One would expect the school to perform better as it had been using Accelerated Reader for a number of years. Mason School of the Laurel District was consistently low. Laurel also has a magnet school. It is possible that the Nora Davis Magnet school could be drawing higher performing students out of Mason, thus lowering Mason’s scores. The magnet school did not score significantly higher than the state average.
Most of the schools in the study have recertified as AR Model or improved to AR Master School in the 2001/2002 school year. Many more schools have also been added (six more in the Pascagoula District). A follow-up study should be done when the 2002 MCT results are released to see if the lower ranked schools came up to the state level and if the others showed improvement. A new study would also have a larger number of schools to analyze.
Cited Sources
Biggers, Deborah. “The argument against Accelerated Reader” Journal of Adolescent & Adult Literacy September 2001 Volume 45 Issue 1. Available on-line: Ebscohost: Academic Search Elite.
Gomez, Dina S. “Reading for the record” NEA Today March 2001 Volume 19 Issue 6. Available on-line: Ebscohost: Academic Search Elite.
Lopez, Steve. “Cat in the Hat and all that” Time 23 October 2000 Volume 156 Issue 17. Available on-line: Ebscohost: Academic Search Elite.
“Texas educators use dual language technology to create successful readers” THE Journal September 1998 Volume 26 Issue 2. Available on-line: Ebscohost: Academic Search Elite.
Topping, Keith & Paul, Terry D. “Computer-assisted assessment of practice at reading: a large scale survey using accelerated reader data”. Reading & Writing Quarterly July /September 1999 Volume 15 Issue 3. Available on-line: Ebscohost: Academic Search Elite.