Dear Dr. Baraka MANJALE Ngussa ,

regarding your question, I can suggest you to have a look at the following note, containing an interesting and detailed answer:

“How to Determine the Validity and Reliability of an Instrument”

By Yue Li

Available at: https://blogs.miamioh.edu/discovery-center/2016/11/how-to-determine-the-validity-and-reliability-of-an-instrument/

Best regards, Pierluigi Traverso.

Thanks Pierluigi for your response. This is so great!

When you purchase some instrument, inquire from the company engineer about its calibration/validation requirements. If you manage that as per the specifications, then you will have the confidence that the results generated from that instrument are valid/reliable.

Thanks Varun for your response. Would you kindly bring your response into the context of research?

Hi Baraka..If you can tell me if you have any specific instrument in your mind, then I can provide a better reply.

Every instrument needs to be calibrated/validated after specific time-duration. for e.g. if you have a auto-pipette for dispensing specific volume of liquid, let's say 1 ml. Then this pipetter eeds to be calibrated after specific interval..mostly annual cycle..to see it's accuracy. Similarly any elctronic thermometer or RH meter also. The more complex the instrument, more intensive is its calibration/validation requirement e.g. HPLC/LC-MS etc.

Thanks Varun. Specifically, I'm talking about research instruments such as quesionnaire and interview schedule

Dear Baraka

I too was misled into thinking about hardware instead of survey “instruments”. Be sure to clarify what you need as much as possible. The previous replies were all very good for hardware instrumentation for analysis. The protocols and methods for evaluation of surveys are much less clear, as so much depends on the language, presentation, and format.

I highly recommend visiting www.surveymonkey.com for excellent tutorials, options, and many free options for simple surveys. Appropriate and valid survey design (as well as execution) are a major research area in themselves. In practice, we try to at least avoid obvious mistakes in interpretation, bias, and analysis. There is no “one best way” to either generate a survey or validate one. There are several common practices that are relevant to all of them. I probably spent more time researching how to create a good survey for some of my doctoral thesis work than I did conducting the survey analysis itself!

As with all research, the most important part is a clear definition of the desired information (not its interpretation or outcome!). Even more so in social/human survey research than in physical science investigations, an ambiguous question – or one with multiple interpretations – will yield even messier and more ambiguous ‘answers’.

Here is another source to begin learning about questionnaire construction:

https://www.projecttopics.org/validity-and-reliability-of-questionnaires-how-to-check.html

Best regards,

Steven