• No results found

3.5 Sample Size and Sampling Techniques

3.5.1 Sample size

Borg and Gall (1989) observes that due to limitations in time, funds and energy, a study can be carried out from a carefully selected sample to represent the entire population. Gay (1992) postulates that for small populations, a sample size of at least 20% of the population is a good representation while for large populations a sample size of 10% is representative enough. This study also used 50% of principals in Nyando district, which is equal to 8 and 50% of science teachers which is 29. To get a representative sample for the students, this study used the following formula;

n=

Where, n is the sample size, N is Population and e is the sampling error/ level of precision (Israel, 2003).

Therefore, using the formulae above, and taking the sampling error to be 0.05, and the student population to be 4500, then, the sample size for this study was 369 students. This means that the study used 8.2% of the students’ population in the district.

29

One principal was selected from boys’ school and one principal from girls’ school in the district while 6 principals were from 14 principals in mixed day schools from district.

6 science teachers were selected from boys’ boarding school, 5 science teachers from the girls’ boarding school and 3 science teachers from each of the 6 mixed day schools sampled. This yielded a total of 29 science teachers for the study

The student sample consisted of 123 from one boy’s school, 90 from one girl’s school and 9 male students and 17 females from each of the 6 mixed day schools sampled. In general, the sample consisted of 369 students, 8 principals, 29 science teachers and 1 DEO. The entire sample matrix constituted 407 respondents. Table 3.1 below shows summary of respondents selected for the study.

Table 3.1: Summary of Respondents Selected for the Study

School category

Respondents Population (N) Sample (n) Percentage (%) Girls boarding Principal 1 1 100

Science teachers 10 5 50 Students 1100 90 8.2 Boys boarding Principal 1 1 100

Science teachers 12 6 50 Students 1500 123 8.2 Mixed day Principal 14 6 42.9

Science teachers 36 18 50 Students Boys 650 54 8.3 Girls 1250 102 8.2

30

3.5.2 Sampling Techniques

Purposive sampling was used to ensure that those selected are the ones who have the specific information needed for the research. In this case, purposive sampling was used to select principals in boys boarding and girls boarding as well as the DEO.

Simple random sampling was used to select the principals from mixed day schools and science teachers from boys’ school, girls’ school and mixed day schools. It was used to get the students from a given school. The simple random sampling ensured that all respondents have equal chances of being selected, thereby ensuring that the sample is representative.

Proportionate sampling was used to select the students in the school categories whereby 8.2% of student population in each school category was picked. Stratified sampling was used to select students based on gender that is male or female. Stratified sampling was used to ensure that there is equal representation across the schools.

3.6 Research Instruments

The research instruments for this study was an interview schedule for DEO, detailed principals questionnaire, science teachers questionnaire and students questionnaire on current performance trends in sciences by gender, current performance in sciences by school category, the strategies adopted by school management to enhance performance in sciences, the strategies adopted by teachers and students of sciences to enhance performance in sciences and the influence of the strategies on school performance in sciences.

31

3.6.1 Interview Schedule

Interview schedule was used because of possible lack of time by the DEO occasioned by busy schedules. They also required short time to gather a lot of information and more information may be obtained from interview schedules. In the interview schedule for DEO; part A sought information on demographic data of DEO and the district, part B on performance in sciences in the district by gender, part C on performance in sciences in the district by school category, part D on School management performance enhancing strategies, part E on teacher related enhancing strategies, part F on student related performance enhancing strategies, part G on influence of performance enhancing strategies on achievements in sciences.

3.6.2 Questionnaires

Questionnaires were used because they are cheap, do not require a lot of effort from the questioner, and they often have standardized answers that make it simple to collate and compile data. Further, questionnaires are an inexpensive way to gather data from a potentially large number of respondents. Often they are the only feasible way to reach a number of reviewers large enough to allow statistical analysis of the results (Mugenda and Mugenda, 2003).

Kothari (2010) recommends multiplicity of the data collection methods noting that the effectiveness of doing so rests on the premise that the weaknesses in each single method is compensated by the counter-balancing strengths of another.

(a) Student’s Questionnaire

In the questionnaire for students; part A sought information on demographic data of student and the school, part B on performance in sciences by gender, part C on

32

performance in sciences by school category, part D on School management related performance enhancing strategies, part E on teacher related performance enhancing strategies, part F on student related performance enhancing strategies, part G on influence of performance enhancing strategies on achievements in sciences.

(b) Science teachers’ Questionnaire

In the questionnaire for science teachers; part A sought information on demographic data of the teacher and school, part B on performance in sciences in the school by gender, part C on performance in sciences by school category, part D on School management related performance enhancing strategies, part E on teacher related performance enhancing strategies, part F on student related performance enhancing strategies, part G on influence of performance enhancing strategies on achievements in sciences.

(c) Principals’ Questionnaire

In the questionnaire for head teachers; part A sought information on demographic data of principal and the school, part B on performance in sciences in the school by gender, part C on performance in sciences by school category, part D on School management related performance enhancing strategies, part E on teacher related performance enhancing strategies, part F on student related performance enhancing strategies, part G on influence of performance enhancing strategies on achievements in sciences.

33

3.7 Pilot study

This involves trying out of research instruments in the field before the actual data collection begins. The piloting of questionnaires was done in one Secondary school during school days. The instruments were presented to respondents similar to the ones who were used in the actual study. The participants in the pilot study did not participate in the final study. Piloting was done to detect deficiencies such as unclear directions, insufficient space to write the response, clustered questions and wrong phrasing of questions. The pilot was able to reveal if the anticipated analytical techniques are appropriate.

3.7.1 Validity

Sherman and Webbs (1997) define validity as the degree to which the participants’ observation achieves what it purposes to discover. Validity of an instrument is the degree to which it measures what it should measure (Coolian, 1994). It is also the degree to which results obtained from and analysis of data actually represent the phenomenon under investigation Orodho (2004). The researcher tested both face and content validity. Face validity is the likelihood that a question will be misunderstood or misinterpreted. According to Wilkinson (1991), pre-testing a survey is a good way to increase the likelihood of face validity. Face validity was tested by considering subjective judgment.

To ensure content validity of the instrument, they were presented to the experts in educational planning in Kenyatta University for scrutiny. The feedback was used to revise the instruments before preparing the final copy.

34

3.7.2 Reliability

The researcher used test-retest technique. According to Mugenda and Mugenda (1999), reliability is a measure of the degree to which a research instrument yields consistent results or data after repeated trials.

Pearson’s Product moment coefficient was employed to compute the correlation coefficient in order to establish the extent to which the contents of the questionnaire are consistent in eliciting the same responses every time the instrument is administered.

The formula for Pearson’s Product moment coefficient is given as: rppm=

Where rppm is Pearson’s coefficient of correlation index, Xi is the ith value of X

variable, Yi is the ith value of Y variable and N is the number of observations or

subjects of X and Y (Orodho, 2003).

Researcher then compared calculated value with critical values from table of critical values of Pearson’s coefficient of correlation; the calculated value was greater than critical therefore a relationship exists between the variables. A correlation coefficient of about 0.75 is normally acceptable. In this study, correlation coefficient of 0.8was high enough to consider the instrument reliable (Gay, 1992).

Related documents