I am doing a reliability analysis of a motivation questionnaire on a sample of athletes in different sports. I do the reliability analysis in order to check the reliability of the translation of the questionnaire into another language.
Thank you.
Alexander Vaculík
Test-retest reliability is one method of estimating reliability. This entails administering the survey to a set of respondents and then repeating the poll to the same group at a later date. The replies at the two timepoints are then compared.
By comparing several copies of the same measurement, reliability may be assessed. It is more difficult to determine validity, although it may be assessed by comparing the results to other relevant data or theory.
The response rate to a survey is usually expressed as a percentage. It is derived by dividing the total number of surveys issued by the number of responders. So, for example, if a survey is issued to 1000 people and 200 of them complete it, the survey response rate is 20%.
The most common way to assess reliability with a single point in time is with coefficient alpha.
Cronbach's alpha coefficient is used for interpreting the reliability of factors derived from either dichotomous or multipoint scales (Abdul Nabi and El-adaway 2021). A Cronbach’s alpha value of 0.75 and above represents a reliable and valid questionnaire (Christmann and Van 2006)
References
Christmann, A., and S. Van Aelst. 2006. “Robust estimation of Cronbach’s alpha.” J. Multivar. Anal. 97 (7): 1660–1674.
Abdul Nabi, M. and El-adaway, I.H., 2021. Understanding the Key Risks Affecting Cost and Schedule Performance of Modular Construction Projects. Journal of Management in Engineering, 37(4), p.04021023.
Best of luck!
Al
Dear Alexander Vaculík first of all try to comply with the back to back translation method to carefully convert a survey from a different language into yours. Then perform pre-test and collect pilot data to better know how your respondents understand the questionnaire. If your pre and pilot testing will produce satisfactory reliability results, the main study sample later will not have any issues. Cronbach alpha reliability and/or Composite Reliability (CR) both are good measures for reliability testing. Tq.
- Badges
- Science method
- Similar topics
- Methodology