Sampling procedures and sample size
This study will be conducted on an 80 student, tutor, and parent population size who are current or past direct or indirect beneficiaries of the peer tutoring program at Stoneybrook middle school. Sample size will be chosen from this population size with a marginal error of 0.5% by the use of snowball non-probability sampling. These chain-referrals forms of sampling procedures are effective when referrals are needed especially when respondents are not easily identifiable and rare for the identification of an unknown potential respondent. This will enable the researchers to identify a few potential respondents and approach them for their participation in the study and ask them for further reference or suggestions of other potential respondents.
The snowball techniques were identified by Etikan, Alkassim, and Abubakar (2016) and are of three types. The first is the linear in which the researcher is referred by the subject to another potential respondent making it a line of referrals. The second is an exponential non-discriminative procedure in which respondents give multiple referrals and all the multiple referrals are explored before going to the next referral level. The other is the exponential sampling procedure in which the researcher selects one referral and leaves the other. In this procedure, the selected respondent refers the researcher to another referral where the researcher also chooses one. This study will use exponential non-discriminative snowball sampling to achieve the required number of respondents. In this study, direct and indirect beneficiaries of peer tutoring programs including students, the tutors themselves, and parents are under study. Upon identifying a key respondent, the researcher will be referred to another potential respondent and the process repeated until the desired established respondent number is achieved.
The established sample size for the study is 80 respondents inclusive of students, tutors, and parents who are directly or indirectly benefiting from the peer tutoring program. The sample size is set at 80 with an allowance of a marginal error of 5% from all over the school composition. The key criteria for inclusion for respondents will be that they have directly or indirectly benefited from their participation in the peer tutoring program. The 80 person sample size will be arrived at using the following formula and procedure;
Formula adapted: Yamane 1967: 886
The sample size was determined using the following procedure,
n = z2pq
e2
n — Desired sample size nless than 10, 000)
z — Standard normal deviate at 1.96 corresponding to 95% confidence interval
P — Estimated proportion of target population that will give a workable sample size with the given level of significance and accuracy.
Q — I — p
e- — The error margin 0.05
n = z2pq
e2
n = 1.962 x 1 x 0.1019 = 79.9711
0.0025
= 80 respondents
Research Instruments
Semi-structured questionnaires will be used as the preferred data collection tool to complement the secondary data sources used in the literature review. A questionnaire and key informant interview will be used. Using more than one instrument helps in supplementing information thus making strong findings and conclusions. The questionnaire will be set to include demographic information of the respondents, the aspects of student tutoring embraced, and the anticipated effects of the engaged approaches (De Vos, Strydom, Fouche & Delport, 2015). A section of opinion about the status of student tutoring will also be included strengthening the part of the recommendations about seeking and venturing into the subject matter.
The qualitative part of data will be collected through the use of a key informant interview guide. The guide will be used to facilitate interviews with selected respondents to support the questionnaire with more information supporting respondents’ choices of answers. The interview guide will mostly include parts included in the questionnaire but leave the prompts in an open-ended approach for the clients to give their own views about the peer tutoring program.
Piloting of research instruments
The instruments will be tested based on a small sample of the population. For the questionnaire, five potential respondents will be selected to answer the questions. From the five, two will also be interviewed to ascertain the applicability of the tool. Piloting of the instrument will be recommended for increasing the validity of the instrument, and ensuring there will be no ambiguous questions. The researchers will also identify questions that need reframing as well as eliminating redundant parts of the questionnaire. The time required to conduct one interview and the time required to fill one questionnaire will be established during the piloting process of the tools.
Testing for validity and reliability/Dependability and Credibility
Validity Issues
After the identification of the right respondents, a pilot study will be conducted at the Stoneybrook middle school with an aim of testing ambiguity, stimulus, consistency, and material of the questions. Researchers will be careful when interviewing stakeholders at a convenient time so that it is done when the respondents are ready as this will serve to maximize internal validity. The researchers will also be careful when making the questionnaire easy to avoid human error and interpretive forgers. Selection of adequate respondent number and other measures based on pretest outside the school with an aim to avoid sensitization and behavior change by participants will ensure validity.
Reliability Issues
A purposeful collection of competent participants through snowballing ensured reliability. Researchers will be careful to avoid bias in compiling data and clearing raw ones, fashioning and entering them into the statistical packages for social scientists (SPSS). For easy and analyzable data format, SPSS version 16 will be used. To ensure consistency in all the procedures, the research assistants will be adequately trained. The result reproducibility in all other settings will be assured by using the standard research protocols, quality control, and universal rules. The periods that the participants will be staying with the questionnaire will also be reduced to minimize the random error. Self-administered questions will ensure that the bias of the interviewer is avoided. The fatigue among the respondents will be minimized by using short single sentence instructions and the questionnaire will be sub-divided into sections. The key informants’ interviews will be made short and direct whereas depth sessions will be direct and lasting five to ten minutes to minimize respondent fatigue.
Data collection procedures
There are two sets of data to be included in the study; primary and secondary data that when triangulated enhance the quality and outcomes of the study. Primary data will be obtained from interviews conducted where two factions are targeted. The interviewees were scholars in this and relevant fields. Structured and unstructured open-ended questions will be set and directed to identified respondents.
Permission will be sought from the school and the department of education for granting permission/letters for the data collection. After identifying the respondents, the researcher and a few selected research assistants engaged in physical will visit the identified respondents. Phone calls and emails will also be sent to potential respondents requesting for the audience to conduct the survey. Once an interview is booked, the researcher will take approximately 10–15 minutes to interview and probe the respondent. Some selected respondents were given questionnaires to answer. The researcher will give an allowance of 2 days to collect the questionnaires for the respondents who cite busy schedules.
Secondary data will, therefore, be extracted from books on treaties, declarations, and policy instruments, previously researched works, journals, newspapers, magazines, and articles written on the subject under study. International formations and policymakers will be considered the core actors and can, therefore, be the greatest source for up-to-date information on the subject matter. The majority of secondary data will be obtained from newspaper articles, journals, periodicals, internet sources, and books.
Data analysis procedures
The research objectives will guide the data analysis process. The collected information will undergo verification before being subjected to the theoretical framework and analyzed through analytical and logical arguments using qualitative approaches. The qualitative information will be analyzed through thematic and content analysis before being integrated/triangulated with the primary data findings. Respondents’ quotes, views, and ideas will be quoted to support the findings and enhance the validity of the results.
For the primary data, information collected through the questionnaires will be subjected to cleaning processes (Etikan, Alkassim & Abubakar, 2016). Data will be coded to make it easy for entry into the Statistical Package for Social Scientists (SPSS) software for processing. Once the data had been entered into the SPSS, the result output will be run leading to the desired output. From the output, the researchers will be able to analyze and interpret the data, while writing the remaining chapters of the study.
Descriptive statistics will be applied to summarize collected data by means of tabulated description (tables), a graphical representation (graphs and charts), and commentary (in-depth narration of the results). For the presentation of data, graphical representations and tables as indicated above will be prioritized to enable the easy presentation of the results
Conclusions
The chain-referral sampling procedures will be used because they can be used to identify respondents who will lead the researcher to others that are unknown but are potential respondents. The researcher will, therefore, be able to identify more respondents, approach them for their participation in the study and ask them for further referral of more respondents until the required sample size is achieved. The study will also adopt a cross-sectional descriptive study due to its significance in facilitating wider data sources coverage. Quantitative and qualitative approaches will be used to provide information on the benefits and experiences of the peer tutoring programs. This is because qualitative studies help give succinct and elaborate answers to research problems (De Vos, Strydom, Fouche & Delport, 2015). The qualitative design will be effective in describing and understanding human behavior while the quantitative approach focuses on the explanation and prediction of human behavior. The use of both approaches, quantitative and qualitative, strengthens the research model as it increases result validity and reliability.
References
De Vos, A., Strydom, H., Fouche, C., &Delport, R..(2015). Research at Grass Roots. Pretoria: Van Schaik Publishers.
Etikan, I., Alkassim, R., & Abubakar, S. (2016). Comparision of Snowball Sampling and Sequential Sampling Technique. Biometrics & Biostatistics International Journal, 3(1), 00055.
