Question

why is a high response in a survey critical? what techniques do social scientist use to increase response rates? (please answer in details)

Answer 1

High response rates are critical in surveys in order to obtain internal and external validity.  Issues here include the need to generalize and to conduct statistical analyses, and the appropriate use of a survey to answer the research question (i.e., methodological validity).

Generalization: Surveys are usually done on a sample of a large population. The sample is designed to represent the population.  If a high percentage of people contacted choose not to respond to the survey, the sample may not represent the population and accurate generalizations of the study results to the population cannot be made from the sample.  Generalizations made under these conditions are likely to be biased or inaccurate.  Analysis: Surveys are usually done on large samples. If we do not get a high response rate there may not be enough cases to conduct adequate analyses, and the study conclusions could be statistically inaccurate or biased.  Methodological validity:  High rejection rates indicate that a) there is a problem with using a survey with this  topic (people aren’t interested in it, or it is too sensitive), b) the survey questions are not written well or are too hard to answer), or c) the method being used to conduct the survey is not appropriate to the topic (for example, using a face-to-face survey in a neighborhood with a high crime rates – people will be reluctant to talk to strangers at their door, let strangers in their homes to conduct the survey).

To obtain high response rates in surveys, we follow the total design method (TDM). Basically this involves strategies that make it easy for the study participants to respond to the survey, such as the following.  1) The researcher should develop valid survey questions and develop a questionnaire with a smooth question ordering.  The researcher should pre-test the survey questions and questionnaire to determine if there are problems with either the questions or questionnaire design/layout.   2) The researcher should develop a mail survey cover letter or a verbal introduction to a telephone or face-to-face survey that accurately and clearly describes the research being done.  This letter or introduction should convey how important the study topic is and how important each selected participant’s response is.  3) With mail surveys, the researcher should include a self-addressed posted envelope for return of the completed survey.  4) The researcher should train survey interviewers (in telephone and face-to-face surveys) to gather the survey data scientifically.  5) The researcher should use multiple attempts to reach participants selected for the study. With a mail survey this involves sending post-card reminders after a week or two of the initial mailing, and an entire new survey packet after 3 weeks.  With a telephone surveys this involves calling selected participants at different days and times. 6) The researcher should consider using alternative survey methods to reach non-respondents – for example, use a telephone call-back to people who haven’t returned a mail survey. 7) The researcher should pilot the survey research design (which includes all of the above) on a small random sample of the population. 8) The researcher should consider using financial incentives for completing the survey such as entering respondents into drawings for prizes, or including a small payment for completing the survey.