Methods Of Sample Selection And The Importance Of Representativeness In Research
Snowball sampling and random sampling as methods of sample selection
The study applied an online recruitment method design of getting the study sample. The sampling technique of the study sample selection is the snowball technique done online too. Qualtrics analysis software was also used to do online analysis on the results (Singh, Farley, & Donahue, 2018). From what is evident in the article this research design method was effective enough as it helped achieve the set goals or objectives of the study including testing the given research hypotheses. The use of snowball sampling also helped achieve acquire participants who could have been hard to reach them. The technique helped avoid time wastage. From my own experience, what could have been an issue in this design is the idea of a high level of biasness during the sample collection (Klumb, & Perrez, 2004). Snowball technique involves using participants to recruit others and this means that in most probable case, participants obtained have the same variables such as age, ease of social media use and when they are friends they can distort everything especially when they do it as friends without being monitored by the researcher. The method was effective overall.
A sample involves the number of participants or the amount, quantity of the item that is applied in the research to help come up with recommendations based on the study outcomes. The right amount, number, quantity of the sample helps determine its appropriateness in getting the desired research goals. The level of appropriateness is also determined by various factors in sampling process which include study population, sample size, sample characteristics and the method of sampling. In research that involves people, a study population involves a group of people sharing some common characteristics selected from a general population to take part in a given study. This population is what gives the researcher a sample (Cleary, Horsfall, & Hayter, 2014). The sample size is determined by the size of population. A bigger population would give a bigger sample size and vice versa. A sample characteristic ought to be the same to make a good conclusion. Characteristics such as age, behavior, height, weight and many other variables need to be the same when comparing the hypotheses of the sample. The method or study design also determines the sample amount to be applied (Lewis, 2015). This method itself may be determined by various factors such as the type of objectives and goals to be achieved. The determination of sample size involves the act of selecting the number of replicates or observations to be included in statistical study sample. The size of the sample is a key feature in any empirical study that involves having the objective of making inferences concerning a given population from the sample. In most practices, the size of the sample is determined based on such factors as data collection expenses and a need to attain a statistical power that is sufficient. Complicated research may have various sample sizes used (Thomas, Buckland, Rexstad, Laake, Strindberg, Hedley, & Burnham, 2010). For example, in a stratified survey the researcher may use different sizes of samples in each stratum. For instance, during the census, the data is collected form the whole population. And this makes the size of the sample and population to be equal.
Factors that determine sample size
In experimental design where the researcher may divide the study into different groups of treatment, different samples may be used for the various groups. There are various ways in selecting sample sizes as follows. The first one is experience which has shown that the choice of small sized samples, though being necessary, may result in confidence intervals that are very wide or high risks of errors during the testing of the statistical hypotheses. The use of target variance to obtain an estimate from a sample that has already been obtained is also another method of selecting a sample in case high precision level is needed. This helps achieve a narrow interval of confidence which translates to a low target of the estimator variance (Lewis, 2015). The other method involves the use of a target for the statistical test power that is applied after selecting the sample. The last method involves the use of a confidence level which translates to the conclusion that the larger the confidence level needed the larger the size of the sample, given that the precision requirement that is constant.
Representativeness in sampling involves selecting a sample that best represents the various characteristics of the population such as age, health condition, or other factors considered to be key during sample selection (Fox, Hunn, & Mathers, 2007). A sample that best represents the population makes it easier to generalize the study outcome and concluding that the results are applicable to the whole population. However this is not the case when the sample is not a representative of the population. In the given article, the population involves university students. The sample size of 124 was quite small to lead to a conclusion and apply the sample results on the whole university fraternity. The size of the sample led to an error due to biasness. It is not clear how the sample represents the population since age is not well factored (McCarthy, Aronson, & Mazouz, 2001). The best sampling technique would have been random method so that the level of biasness is controlled. It would also have increased the level of sample representativeness and thus applying the study results on the university population would not have any issue.
In the article, the measurement of behaviors related to the social networking sites involved identifying the frequency of posting selfies, perceived level of attractiveness of these selfies, and the frequency of likes and tags on the selfies, on balanced scales of rarely, multiple times in a day, below and above average, too much or too little (Singh, Farley, & Donahue, 2018). Participants were also required to record the number of hours per week each spend on the social sites. Though these measures were done well, there contribution to good results may not be the case. The reason is that the behavioral variables were close-ended and the participants were only supposed to select the best choice among the many given. Item validity and high response accuracy may involve the use of such tools as an open-ended type of a questionnaire. This tool needs to accompany the closed-ended type of questionnaire so that the participants can be able to give a little explanation as to why and when of the social site usage. The other method involves the use of a different method design of data collection which may be an interview. This could mean conducting a live interview or on the phone (Klumb, & Perrez, 2004). The method could ensure that the data that is being obtained has some high level of validity or accuracy in responses.
Sometimes interviews may prompt the participant to give the correct answers due to the fear of telling a lie. The researcher can also probe for more answers the moment they sense a need for clarification. The other method may involve conducting the study as a focus group which involves a controlled and regulated discussion among participants. They can raise queries in case they sense that there level of response and item validity is low prompting to further discussion to reach a consensus. However, there are various factors that prompt for the use of a given method such as the one applied in the study. These include the challenge in getting the participants and hence the researchers settled for the online close-ended questionnaire (Singh, Farley, & Donahue, 2018). The suggested methods could have been applicable based on aspect of time, sample availability, cost of conducting the study and the need to test the use of the online method in carrying out research. In this case, it can be said that validity of the items used helped achieve the desired goals even when the level is not quite high.
References
Cleary, M., Horsfall, J., & Hayter, M. (2014). Data collection and sampling in qualitative research: does size matter?. Journal of advanced nursing, 70(3), 473-475.
Fox, N., Hunn, A., & Mathers, N. (2007). Sampling and sample size calculation. The NIHR RDS for the East Midlands/Yorkshire & the Humber.
Gliner, J. A., Morgan, G. A., & Leech, N. L. (2011). Research methods in applied settings: An integrated approach to design and analysis. Routledge.
Klumb, P. L., & Perrez, M. (2004). Why time-sampling studies can enrich work–leisure research. Social Indicators Research, 67(1-2), 1-10.
Lewis, S. (2015). Qualitative inquiry and research design: Choosing among five approaches. Health promotion practice, 16(4), 473-475.
McCarthy, R., Aronson, J., & Mazouz, K. (2001). Measuring the validity of task technology fit for knowledge management systems. AMCIS 2001 Proceedings, 337.
Singh, S., Farley, S. D., & Donahue, J. J. (2018). Grandiosity on display: Social media behaviors and dimensions of narcissism. Personality and Individual Differences, 134, 308-313.
Thomas, L., Buckland, S. T., Rexstad, E. A., Laake, J. L., Strindberg, S., Hedley, S. L., … & Burnham, K. P. (2010). Distance software: design and analysis of distance sampling surveys for estimating population size. Journal of Applied Ecology, 47(1), 5-14.