During research, it is of great importance for the researcher to discuss the degree of validity and reliability of measuring instrument in research.

To improve the reliability of data collection instrument, the researcher has to do the following:

1. To adopt a test and re-test administration, if the researcher is using questionnaire. This involves administering the same questionnaire to same group of respondents at different point in time also known as pilot test.
2. Another method of establishing reliability is to ask questions that increases the chances of memory relapse, failure or loss.
3. The question should not be ambiguous
4. Clear and standard data collection instruments should be provided to forestall error.

To improve the validity of an instrument, the following should be done;

1. An expert opinion should be sought to ascertain validity of the instrument. In most cases the research supervisor is best fit to determine the validity of the research instrument.
2. The content validation of the research instrument is essentially based on one’s judgement. In an attempt to validate the instrument, a questionnaire could be administered to selected respondents who may not be included in the sample with aim of determining whether the questions will mean same thing to the entire respondents.
3. Criterion related validity could be used by comparing tests or scale scores with one or more external variables or criteria known or believed to measure the attribute being studied. For instance, criterion related validity can be used to predict the success or failure of students from a university’s Post UTME exam. The problem with criterion related validity is that it could be difficult to achieve the external criterion with which to draw comparison.
4. The use of concurrent validity as a measure for calculation of correlation coefficient between the distribution of test scores and some concurrently existing criterion measure.

Use of predictive validity to measure data or prediction of future events. E.g., to determine the performance of students in a particular exam, predictive validity will be obtained by computing the correlation coefficient between a distribution of test scores obtained at an earlier time against a distribution of scores on some later criterion measure.