177

The Effects of Computer Assisted Instruction on StudentAchievement in High School Biology

Patricia D. Morrell Department of Science EducationOregon State UniversityCorvallis, Oregon 97331

Computers first found their way into the classroom in theearly 1960s. In less than three decades, computers havebecomeas commonplaceinmostschool districts asbooks. Consequently,computer literacy and programming courses are widespreadandpopularcourse selections. Unfortunately, useofcomputersby subject area teachers at the middle and high school levels isnot a typical practice. Bangert-Drowns, Kulik, and Kulik(1985) reported that rather than being used as a teachingsupplement, more than 85% of computer use is limited tocomputereducation classes--a statementsupportedbyMcManes,Cannings, and McCall (1985).

Among the many reasons why computers are not widelyused as an educational tool are that good quality software isscarceand much software is limited to run on only onecomputersystem (Forman, 1982; Summerlin & Gardner, 1973; Tocci,1981). Additionally, school curricula have notbeen redesignedfor effective use ofcomputers (Mojowski, 1987). Finally, littleinformation is available on how effective computers are whenused as teaching tools (Bangert-Drowns et al., 1985; Forman,1982; Ybarrondo. 1984).

The effectiveness of computer drill-and-practice programson basic skills attainment on the enhancement of classroomactivities by computer simulations has been the topic of manyprevious investigations (Bangert-Drownsetal., 1985; Edwards,Norton. Taylor, Weiss, & Duseldorp, 1975; Farthing, 1975;Forman, 1982; Gallagher, 1987; Hardin County Board ofEducation, 1984; Kulik, Bangert, & Williams, 1983; Nakhlch,1983; Smith, 1984; Summerlin & Gardner, 1973). Few studieshave been reported that deal with the computer’s effectivenessas atutor, aroleforwhich thecomputeris well-suited (Summerlin& Gardner, 1973; Ybarrondo. 1984).

If one looks for information just on science tutorials insecondary schools, the research is even more scarce. Ofthe 500titles examined by Bangert-Drowns et al. (1985), only 42 werefound to be what they considered well-planned and conductedresearch. Of these, only one was in high school science.Edwards et al. (1975) listed no high school science tutorials intheir review. Summerlin and Gardner (1973), Ybarrondo(1984), and Gallagher (1987) all stated CA1 in science appearsto be an inadequately tested area.

Purpose

There were two main objectives to this study:1. To determine if there is a significant difference (a = .05)

between the mean posttest achievement scores of studentsreceiving CAI instruction and students receiving lecture/discussion forauniton photosynthesis and auniton introductorygenetics.

2. To assess student attitudes toward CAI.

Design

Three classes of general biology students at a rural highschool in Oregon were used in this study. This sample waschosen for convenience. The students in each class period werelisted alphabetically in decreasing orderbased on the average oftheir first and second term biology letter grade. The lists werethen alternately separated into the treatment group and controlgroup, producing three sets of two groups of comparableachievement in the subject matter.

The units chosen for this study were photosynthesis andintroductory genetics. These two areas were selected becausestudents find them troublesome, the content can be easilyadapted to a tutorial, the introductory materials can bepresentedin a short period of time, and the topics lend themselves toobjective testing. For the photosynthesis lesson, 27 studentswere in the control group and 29 were in the treatment group;the genetics lesson had 27 in the control group and 27 in thetreatment group.

For each unit, objectives were listed. Then a complete set ofdetailed teaching plans was written to meet these objectives.Next, a tutorial-type CAI was developed in the ApplesoftBASIC language based on the teaching plans and covering thesame objectives. The CAIs were prepared following theguidelines suggested for effective software development(Klopfer et al., 1983; Nakhleh, 1983; Vargas, 1986). Thesoftware included graphics and text and was user-paced. Self-test quizzes were included, which provided brief explanationsfor incorrect responses. User involvement was stressed. Astudent manual was also developed to accompany the softwarefor each unit.

For the photosynthesis unit, a 20-question multiple choiceposttest was written to test the objectives; for genetics, theposttest consisted of 18 multiple choice questions, 4 matchingquestions, and the construction ofa Punnett square for a simplecross. The questions for both tests were a mixture of originalqueslionsondsomQfromResourceBookofTestItems,BiologicalScience: An Ecological Approach, BSCS Green Version, 5thEdition (Meyer & Meyer, 1982). A university computer

Volume 92 (4), April 1992

Computer Assisted Instruction

178

science educator compared the teaching plans with the tutorialsoftware to ensure the two methods paralleled each other andmetthe stated objectives. She alsoexamined the tests forcontentvalidity. The Kuder-Richardson formula 21 reliabilitycoefficients for the photosynthesis test was 0.66 and for geneticswas 0.72. These scores require the resultant achievement datato be viewed cautiously, although they are sufficiently high formean comparisons.

The photosynthesis unit was completed in fourclass periods;the genetics unit was completed in three. The length of a classperiod was 49 minutes.A survey was written to assess student attitudes toward CAI.

The instrument was comprised of 25 questions to be answeredon a scale of 5, choices ranging from strongly agree to stronglydisagree. To ensure the tone of the question would not affect astudent’s response, most of the statements were included in twoplaces on the questionnaire, written once favoring CAI andagain in unfavorable wording. The user was also asked forcomments and/or suggestions. Signing the form was optional.

Treatment

The students were not told that they were to be part ofa studybut rather that they were going to try something new� some ofthem would be taught by computer while the others wouldremain in class. Later in the term, they would switch so thosestaying in the classroom now would work with computers laterand vice versa; that is, the control group and treatment groupswould reverse roles. Most students were enthusiastic about theopportunity to work with the computers; all were willing toparticipate.

The students were divided into the pro-determined groups.Each students in the treatment group received his or her owncopy of the student manual. Aside from reporting to class forattendance, there was no instructional contact with their regularbiology class during the tutorial period. After checking in, thetreatment group went directly to the computerroom and workedwith the tutorial CAI until the end of each class session. Thesestudents were directed to work at their own pace through theprogram, take notes in the accompanying manual, and go overany sections as often as they wished. Generally, the studentsworked independently but occasional hardware problemsrequired students to pair up for parts of the tutorials. Whenstudents completed the tutorial, they were given the attitudesurvey to complete. The students did not rejoin the regular classuntil the end of the 3-day (genetics) or 4-day (photosynthesis)scheduled time frame.

The control group met in the regular classroom. Teachingwas performed in the usual informal lecture/discussion style.The blackboard was used as a visual teaching tool. Theaforementioned teachingplans were carefully followed so as notto deviate from the prescribed instruction.

At the close ofeach ofthe two units, thegroups werereunitedand the posttest was administered. There was a span of two

months between the two selected CAI units.

Analysis

Theposttests were graded and the students’ test scores wereentered into the Number Cruncher Statistical System, Version5.01 (Kaysville, Utah) for statistical analysis.

To help in analyzing the student attitude questionnaire, thefive categories ofstudentresponses on the survey were assignednumerical values: strongly agree = 5 points, agree = 4 points,undecided = 3 points, disagree= 2 points, and strongly disagree= 1 point. The questionnaires from the two treatment groupswere analyzed separately. Average response values werecalculatedby multiplying the numberofstudents responding ineach particular response category by the numerical valueassigned to that category. The products were added anddivided by the number of respondents. Due to the non-cumulative nature of the survey instrument, the responses toeach question by the two treatment groups were statisticallyanalyzed question by question, by means of a Most. Since theconclusions of this investigation were not contingent upon theassumption that no Type 1 errors existed among the statisticaltests, the problems normally associated with multiplesignificance tests were not a concern; however, because as oneincreases thenumberofnonindependentsignificance tests, thechances of finding significant results increases, the actual p-values for each test item was provided.

Results and Discussion

The null hypothesis tested for each of the posttest analyseswas that no significant difference at the 0.05 level would befound between the means of the posttests of students in thecontrol and treatment groups.

The computer-generated statistical data on studentperformance on the posttest for the photosynthesis unit aresummarized in Table 1. From the table, it can be seen that nosignificantdifferencewas foundbetween the means ofposttestsof students receiving the two treatments. As a result, the nullhypothesis cannot be rejected.

Table 2 summarizes the computer-generated statistical dataon student performance on the posttest for the genetics unit.Again, the results failed to reject the null hypothesis.A paired r-test was used to compare the scores of the

students who participated in the photosynthesis CAI andtraditional genetics lessons. The students’ posttest scores wereconverted from raw points to percentages for purposes of thiscomparison. Table 3 summarizes these data. There was nosignificant difference between the means of these posttestscores, and the null hypothesis cannot be rejected.

Table 4 summarizes the data of the paired /-test for thescores ofthe students who participated in the genetics CAI andtraditional photosynthesis lesson. The datado showadifferencebetween the means of the posttest scores which is statistically

School Science and Mathematics

Computer Assisted Instruction

179

Table 1

Two Sample t-test Results of Pastiest Achievement Scores ofStudents Completing the Traditional and CAI PhotosynthesisUnits

Number in SampleMean1Standard Deviation/-value

Probability

Traditional

2714.703.73

-0.5941.5549

CAI

2914.172.94

�Maximum score = 20 points

Table 2

Two Sample t-test Results of Pastiest Achievement Scores ofStudents Completing the Traditional and CAI Genetics Units

Number in SampleMean1Standard Deviation/-valueProbability

Traditional

2717.223.90

-1.6326.1086

CAI

2715.334.57

a Maximum score = 25 points

Table 3

Paired t-test Results of Pastiest Achievement Scores of theSame Students Receiving Different Treatments for TwoUnits: CAI Photosynthesis Unit & Traditional Genetics Unit

Genetics -Traditional Photosynthesis-CAl

Number in Sample 2727Mean1 68.8971.11Mean Difference2.22Standard Error ofDifference2.8568/-value.7771Probability.4441

�Maximum score =100 points

significant. The null hypothesis is rejected.Finally, a paired /-test was used to compare each student’s

achievement using the two treatments. The statistical data aresummarized in Table 5. The null hypothesis cannot be rejectedas no significant difference exists between the posttest means.

Although the data from the genetics unit showed treatmentsby the two methods were not significantly different, the data inTable4 questionswhy the studentsexposed to the geneticsCAIdid significantly more poorly in this unit than in the traditional

Table 4

Paired (-test Results of Pastiest Achievement Scores of theSame Students Receiving Different Treatments for TwoUnits: CAI Genetics Unit &. Traditional Photosynthesis Unit

Photosynthesis-Traditional Gcnetics-CAl

Number in Sample 2727Mean1 72.4161.41Mean Difference11.00Standard Error of Difference3.2007/-value3.4367Probability.0020

�Maximum score =100 points

Table 5

Paired t-test Results of Posttest Achievement Scores of theSame Students Comparing Traditional Treatments and CAJ

Traditional CAI

Number in Sample54 54Mean*70.46 66.26Mean Difference-4.20Standard Error ofDifference2.3150/-value-1.8158

Probability.0751

�Maximum score =100 points

photosynthesis lesson. There may be several possible reasonsto explain this. The genetics unit had many new terms andabstract concepts. For the most part, students were not familiarwith the material presented in this unit. Conversely, thephotosynthesis unit had few new vocabulary words and wasmore of an enrichment lesson, building on prior concepts andextending what the students learned about photosynthesis in thelower grades. Compared with photosynthesis, the genetics unitwas of a higher level in Bloom’s taxonomy. Accordingly, theposttest scores were lower for this unit than the photosynthesislessons, regardless of treatment. The genetics unit requiredmore abstract thought, deductive reasoning, and application.

In conclusion, it would appear the CAI tutorials were nomore or less effective in promoting student achievement thanthe traditional lecture/discussion approach.

Thestudentresponses to the questionnaire measuring studentattitude toward CAI were tabulated and analyzed as explainedpreviously. Theaverage response value,/-value, and probabilityvalue for each item on the questionnaire are presented in Table6. The analyses show that there were definite differences instudent attitude toward CAI depending on which CAI unit thestudent worked with. In general, few students from eithergroupfound CAI units to be too mechanical or complicated to use.

Volume 92 (4). April 1992

Computer Assisted Instruction

180

Table 6

Average Response Value, t-values, and Probability Values/or Each Item on the Student Attitude Questionnaire Completed byStudents Participating in CAI Units"

Statement

No one cared if I learnedMore involved with computer than lessonWorked at own paceCAI is too mechanical*Uncomfortable with CAICAI is time efficientCAI is more motivating than classroom instructionCAI makes the subject more interestingCAI is too impersonalPrefer classroom teachingDidn’t care if missed a questionLiked the material before CAI*Liked the material after CAI*Want more CAI unitsPressured to hurry upComputer is not distractingCAI is impersonalCAI makes the material less interestingComfortable with CAI*CAI is inefficient timewiseClassroom teaching is more motivating than CAIPrefer CAIDidn’t like the material before CAIDidn’t like the material after CAI*Do not want more CAI units

Photo

1.922.284.202.212.084.043.843.922.322.322.123.163.714.042.603.882.802.604.082.562.323.843.042.582.04

Genetics

2.553.053.643.183.322.772.912.683.363.822.413.432.952.773.273.233.453.363.002.723.862.142.643.183.54

(-value

-2.416-2.3362.023-3.183-3.3814.6672.1903.246-4.334-3.780-0.938-0.9963.1313.157-1.8402.245-3.101-1.9963.382-0.520-4.0624.6091.464-2.493-3.482

Probability

.020

.025

.050

.003

.002

.000

.034

.002

.000

.000

.353

.325

.004

.003

.072

.030

.003

.052

.002

.606

.000

.000

.150

.016

.001

b

"Forty-seven students completed the questionnaire, except for the question marked with an asterisk which had 46 responses.

^The response categories available to the repondents were SA = strongly agree, A = agree, UN = undecided, D = disagree,SD = strongly disagree. Numerical values were assigned to the response categories, with SA = 5 points, A = 4 points, etc. downtoSD= 1 point. These values were multiplied by the number ofactual responses in that category for each statement. Theresukingproducts were added and divided by the number of respondents for that particular item.

They did not feel as though no one cared about whether or not latter disagreed.they learned and were themselves concerned with selecting the More revealing were the results that the photosynthesiscorrect responses when prompted by the computer, group did not favor traditional classroom instruction over CAI

The photosynthesis group felt the treatment was an efficient and wanted more CAI units. Conversely, the genetics groupuse of their time, not impersonal, and was comfortable with the did not prefer CAI units and did not want more units taught intreatment. In contrast, the genetics group disagreed their time this manner.was used efficiently and was less certain as to how impersonal Based on the results ofthe questionnaires and the additionalthe treatment was and how comfortable they felt with the comments provided by some students, it would appear thattreatment, those in thephotosynthesis unit weremore favorable to learning

The photosynthesis group found CAI to be more motivating with a computerized tutorial than those students working withthan classroom instruction while the genetics group found CAI the genetics unit. Given the higher mean posttest scores on theto be less so. Additionally, the first group leaned toward the photosynthesis test, one has to wonder whether the attitude offeeling that CAI made the material more interesting while the thestudentsaffectedtheirresultsorwhetherthephotosynthesis

School Science and Mathematics

Computer Assisted Instruction

material was indeed easier toleam making the entireexperiencea more favorable one for those students involved.

Recommendations

It is intuitively obvious that more studies need to be done todetermine the effects of CAI tutorials on student achievementin high school science. This study, like most others, did notoffer much light on the academic worth of using CAI tutorialsin classrooms.

There are some recommendations which can be madeconcerning replication ofthis particular study. First, the studentsample needs to be increased. A wider base might offer moreinteresting, revealing data. Second, involvement of severalteachers would improve any replication. Student academicachievementcanbeinfluencedby theteacherofthe instructionalmaterial. Since only one teacher was used in this study, it maybe that the academic achievement of students receiving thelecture/discussion treatment was more of a reflection of oneindividual’s teaching skills than of the method itself. If severalteachers lecture using the prepared materials, the instructionalbias could be investigated as a possible variable or removed.Additionally, having the students work in groups oftwo or threeon the computers could serve to alleviate the negative aspectsof independent work. The biggest complaint of the CAIstudents was the inability to ask questions. If they wereworking consistently in groups, perhaps the group couldcollectively deal with problems as they arose.

The inability of students to interact with the teacher pointsto a major area ofweakness in this and similar studies. A moregeneralized design wouldprovidestudents with theopportunityto get teacher assistance as they work on the computer. In thisinvestigation, it was not possible for the teacher to interact withthe students. The effectiveness of using a teacher/CAl tutorialcoupling versus traditional lecture/discussion methods is anarea that has not been, and should be, more fully investigated.A main drawback to the use of computers in science

classrooms today is the lack of available quality software. Ifsoftware were written following the guidelines research hasshown to be effective, there would be less apprehensionconcerning the incorporation of computerized lessons into theschools. The question of the effective of CAI instruction onstudent academic achievement in high school biology stillremains unanswered; however, due to the positive attitudesstudents generally have toward working with CAI units, it is anarea that merits further research.

References

Bangert-Drowns. R. L., Kulik, J. A.. & Kulik, C. C. (1985).Effectiveness of computer-based education in secondary

181

schools. Journal ofComputer-BasedInstruction, 72(3), 59-68.

Edwards. J.. Norton. S.. Taylor, S.. Weiss. M., & Dusseldorp,R.(I975). How effective is CAI? A review of the research.Educational Leadership, 33, 147-153.

Farthing, F. (1975). Computer-assisted instruction: Somecurrent literature. In Department of Computer Science(Ed.), Computers in Educational Resource Handbook,Eugene: University of Oregon.

Forman, D. (1982). Search of the literature. The ComputingTeacher, 9(5). 37-51.

Gallagher, J. J. (1987). A summary of research in scienceeducation-1985. Science Education, 71 (3), 358-364.

Hardin County Board ofEducation. (1984). The fourth basic:Computer skills (Final Report). (ERIC DocumentReproduction Service No. ED 259 146)

Klopfer. L. E.. Abegg, G. L.. Batoff, M. E., Doyle. J., Finley, F.N.. Horak, W., Keane, J., Luncsford, D., & Lunetta, V. N.(1983). Microcomputer software evaluation instrument.Version 1983. The Science Teacher, 57(1), 95-98.

Kulik, J. A.. Bangert. R. L., & Williams. G. W. (1983). Effectsof computer-based teaching on secondary school students.Journal of Educational Psychology, 75(1), 19-26.

McManes. J.. Cannings, T.. & McCall, C. (1985). Developinginstructional applications at the secondary level: Thecomputer as a tool. (ERIC DocumentReproduction ServiceNo. ED 265 850)

Meyer. D. E., & Meyer. W. V. (1982). Resource book of testitems, biological science: An ecological approach, BSCSgreen version (5th Ed.). Boston, MA: Houghton MifflinCo.

Mojowski, C. (1987). Technology and curriculum: Will thepromised revolution take place? NASSP Bulletin, 77(496),113-118.

Nakhleh, M. B. (1983). An overview ofmicrocomputers in thesecondary science curriculum. The Journal ofComputers inMathematics and Science Teaching, 3(1), 13-21.

Smith, S. G. (1984). Computer-assisted instruction on amicrocomputer. Journal of Chemical Education, 67(10),864-866.

Summerlin, L. & Gardner, M. (1973). A study of tutorial-typecomputer assisted instruction in high school chemistry.Journal of Research in Science Teaching, 70(1), 75-82.

Tocci.S.(1981). The microcomputer/biology "interface." TheScience Teacher, 45(5). 60-62.

Vargas, J. S. (1986). Instructional design flaws in computer-assisted instruction. Phi Delta Kappan, 67(10). 738-744.

Ybarrondo, B. A. (1984). A study of the effectiveness ofcomputer-assisted instruction in the high school biologyclassroom. (ERIC Document Reproduction Service No.ED 265 015)

Volume 92 (4), April 1992