In any research endeavor, the researcher strives to use the smallest possible sample, including all the population traits. In that case, the choice of a sampling method is critical because it dictates the amount and quality of representation of the entire population. Originally, sample frames were primarily composed of lists of email addresses collected for other purposes. Today, the initial panel recruitment is often done by phone or email, and databases or panels are "rich" sampling frames. The samples can include personal data on respondents collected during their recruitment. In the Szolnokin and Hoffmann (2013) case, the authors delve deep into this issue and assess its impact on the result by comparing online, telephone, and face-to-face surveys as important sampling methods. However, the combination of different collection methods poses new challenges in data quality and analysis (Sarstedt et al., 2018).
In Szolnokin and Hoffmann (2013), the comparison of telephone and internet surveys and face-to-face sampling methods in the wine consumption. The assessment looks into each group's dynamics and assesses which is the best and more applicable in research. Of course, each of these methods poses challenges and problems and elicits differences in respondents' cognitive processing of questions. An interviewer's presence or absence often affects the answers to sensitive, difficult questions or may call for socially desirable responses. When the questions have many possible answers, respondents make better and precise choices when addressing questions or a list of items compared to when a surveyor reads the same to them. Thus, the authors make frequent observations that respondents by telephone give more socially desirable answers and more often choose the items that they were read last. According to Gong et al., (2019), responses are more concentrated at the extremes in some cases and for many responses, which is a slight indication that a face-to-face survey is inherently different from both telephone and online research. These phenomena can be referred to as "effects of modes of administration" of questionnaires impact research and negatively affect data quality in a research project.
Delegate your assignment to our experts and they will do the rest.
In the latter case, the sub-samples of the three collection modes may differ on key characteristics. Another common cause of mixing is that in which the sub-samples belong to separate bases that partially overlap. Here, the expectation is that the frame coverage and the non-response bias will differ between the two collection modes. The modes again result in discrepancies in the distribution of responses in the different sub-samples. In simple comparisons of the results of three subsamples, the effects of administration and selection are confused. For example, the respondents in an internet sub-sample are more likely to declare risky behaviors. Such dynamic has an administrative effect in as far as such behaviors are concerned because it becomes much easier to admit on the internet in the absence of an interviewer, and it drastically improves the quality of data because it improves accuracy.
On the contrary, it undermines the selection effect that is only prevalent if the respondents' age distributions are different in the three sub-samples or live in other regions. As a recommendation, the authors should have endeavored to improve the quality of data by including many respondents in different regions of the country, using an online survey. In more complex comparisons, controlling for variables associated with selection effects may be much complex, and the remaining differences are likely due to administration effects. The administration effects can then be modeled when the telephone and internet subsamples are combined.
In conclusion, Szolnokin and Hoffmann (2013) show how sampling dictates surveys, models and determines data quality. However, the choice of sampling method may be tweaked based on the respondents. In psychology, it is evident that respondents seem to admit more when doing online surveys as opposed to face-to-face surveys. In that case, the recommendation is to combine all three survey methods to improve data quality instead of relying on a single methodology in the sampling survey.
References
Gong, W., Taighoon Shah, M., Firdous, S., Jarrett, B. A., Moulton, L. H., Moss, W. J., ... & Chandir, S. (2019). Comparison of three rapid household survey sampling methods for vaccination coverage assessment in a peri-urban setting in Pakistan. International journal of epidemiology , 48 (2), 583-595.
Sarstedt, M., Bengart, P., Shaltoni, A. M., & Lehmann, S. (2018). The use of sampling methods in advertising research: A gap between theory and practice. International Journal of Advertising , 37 (4), 650-663.
Szolnoki, G., & Hoffmann, D. (2013). Online, face-to-face and telephone surveys—Comparing different sampling methods in wine consumer research. Wine Economics and Policy , 2 (2), 57-66.
