Reliability and validity concepts are the main issues used to determine the value of survey information. The act of determining the quality of survey information is a technical activity that is vital in evaluating the sufficiency of the data to aid audit observations. For a survey to provide sound, relevant, sufficient, and consistent evidence, the data it relays must be valid and reliable (Kimberlin & Winterstein, 2008). The data are reliable if the measurement is consistent from the individuals surveyed, at different times, and across settings. Having consistent information is vital in the making of general statements. Validity on the other hand is the degree to which the data obtained from the survey is significant in making a conclusion (McCrae et al. 2011). Validity emphasizes on the accuracy of the measurement, and it is deliberated in a representative sample context. Validity is impacted by the survey design as it depends on asking questions to measure what is supposed to be measured (Kimberlin & Winterstein, 2008). This paper aims to discuss the steps and types of validity and reliability to be used in identifying the teenagers who drive while intoxicated.
Types of Validity and Reliability
According to studies, most surveys have face validity. Apparently, asking questions is deemed as the best way to get the information being searched , but a question arises are they really? For this reason, writers should use other forms of validity which include;
Content validity is linked to the ability to come up with questions reflecting the problem at hand while ensuring that the most relevant subjects are not left out. For instance, in this case the questions will not only be aimed at getting information on the teenagers at risk of driving under the influence but rather the implications of driving while intoxicated.
Internal validity focuses on whether the questions posed can actually explicate the results required for the survey (McCrae et al. 2011). The relationship between independent variables and the dependent variable is put into consideration.
External validity which is the degree to which the survey outcomes can be generalized to the whole population represented by the sample (Milton et al. 2011). It is true that how questions are asked has a great impact on the answers obtained and as a result, the questions must mirror the perceptions of the target population regarding the research problem.
Test- retest reliability obtained by performing the same test to a group of people two times. The scores obtained from the two tests can be linked to assessing stability test over time.
Parallel forms reliability which is obtained by performing distinct versions of an evaluation tool for the same group. The two versions should contain items that investigate similar construct, knowledge base and skill.
Inter – rater reliability is used to evaluate the extent to which various judges agree on the evaluation decisions. This measure of reliability is important as the human observers will not interpret the feedback the same way.
Reliability necessitates the employment of consistent information collection tools and survey processes that are intended to improve consistency. To obtain relevant outcomes, careful planning as well as obtaining information from the best suited persons should be considered (Marsden & Wright, 2010). Obtaining accurate and complete information necessitates excellent survey administration processes.
Steps to establish reliability and validity of the survey
In order to establish the reliability and validity of a survey, the following steps should be employed:
- The survey should encompass diverse sets of questions that measure similar attribute. This is termed as the internal consistency method of establishing reliability.
- Factors that may affect the survey’s internal validity must be put into consideration. To ensure a good internal validity, the questions should be clear and appropriately worded.
- The survey sample should be representative of the general population. Random sampling should be used and the sample size must be large enough. A large sample size is recommended as it improves the accuracy of the results and lessens the ‘margin of error’.
The act of determining the quality of survey information is a technical activity that is vital in evaluating the sufficiency of the data to aid audit observations. Reliability and validity concepts are the main issues used to determine the value of survey information. The paper discusses the steps and types of validity and reliability to be used in identifying the teenagers who drive while intoxicated.