The pandemic has resulted in the submission of numerous self-developed nonvalidated questionnaire-based manuscripts by authors. These nonvalidated questionnaires may deviate from the outcome and provide a false interpretation of the described study objective. A questionnaire for survey-based research contains a set of questions that are developed to collect data related to demography, health-related information, or opinions from respondents.
The development of a new questionnaire is nonessential when a validated questionnaire is available in the literature. Authors should have a thorough knowledge of framing a questionnaire if a void exists in the literature. It needs to be reinforced that a questionnaire in the English language must be validated when administered in a regional language to the regional participant. For example, an oral health impact profile-14 questionnaire administered in a regional language to an individual would alter the meaning or intended objective of the questionnaire. The author should also consider performing both forward (from English to Regional language) and backward (from translated Regional language to English) translations. The forward translation of the regional language is reliable only if the backward translation provides the same meaning as the original questionnaire. Blinding the backward translators about the intended idea/objective of the research would prevent bias. Apart from language, the questionnaire that has been intended for a particular age, sex, or social group will not match the other subdivision of the group, and hence these also require revalidation based on the objective of the study design. Authors should understand that exhibiting their high proficiency in the language while framing questions and lack of training with the administrator, especially in interview-based questions, makes the participant misunderstand the actual meaning of the questions.
Development/validation of a new questionnaire begins with a literature review to identify the void that necessitates the framing of questions. The essential aspect of framing a survey-based research questionnaire is shown in Figure 1.
An observation or interview with a few participants by the focus group method helps identify the problem from the population's perspective. Literature search being an imminent part of any research, obtaining an expert interview specialized in the field is an essential aspect before framing questions. Based on focus group discussion and expert opinion, the researcher should synthesize the questions correlating to his/her objective; each objective becomes a domain with a set of questionnaires.
Validation of each question begins with a preliminary cognitive interview of 10 participants and checking the participant's way of understanding the question. The question should be selected based on the analogous outcome of these participants, followed by the removal of unwanted questions, or modifying the language of the question. After cutting the clutter, the researcher should redo the cognitive interview with another set of 20 participants in two sets, with 10 in each, to confirm cognitive validation.
The reliability of the questionnaire follows cognitive interview and should involve checking with at least 30 participants. It is judged by the reproducibility of similar answers when the questionnaire is administered at different intervals of time, with a minimum of 15 days gap between the interviews. However, it is difficult to evaluate the reliability of pain at different intervals since the patient experiences pain at different levels with time. Reliability also helps to determine errors present in content sampling, variations in demographic characteristics of respondents, measurement scales, etc. There are multiple aspects in evaluating reliability; internal consistency, test–retest reliability, inter-rater reliability, parallel form reliability, and split-half reliability. Internal consistency, measured by Cronbach alpha co-efficient that evaluates the inter-correlation of items in the questionnaire, and the reliability of 0.7 is adequate to check the validity to a wider population. Test-retest reliability is administering the questionnaire to the same respondents at a different period of research and measuring the Pearson correlation. Parallel form reliability is the evaluation of two different domains of a questionnaire by the same participant, responding to the first domain followed by the second domain. Split-half reliability is the evaluation of two different domains of a questionnaire by splitting the participants and providing the questions of each domain to each of the groups simultaneously. We also need to check the Inter-rater reliability between different questionnaire administrators and a kappa coefficient above 0.61 is an acceptable agreement. The researcher should understand that it is an unethical practice to increase the number of questions to increase the reliability. A participant will ignore the “n“ number of questions and responds without even reading the questions, especially in online mode.
Validation is done after reliability to confirm the constructed questions measure the intended objective of the study. For example., assessing the oral health quality of life specific to cardiac disease measures the output of oral health from cardiac disease and not because of their social status or health issues. Validity also has several phases: face validity, content validity, construct validity, criterion validity, and convergent validity. Face validity is the linguistic analysis, and the response from each person depends on the IQ or understanding ability of the respondent. The response also depends on their age, educational, and socioeconomic status and is assessed by the Cohens kappa index. Content validity is the expertise given by the panelist to evaluate the efficiency of the chosen items to measure the variables in a domain measured by Lawshe's Method. Construct validity determines the set/sequence of a question under each domain that correlates with the objective. It is considered small if the correlation coefficient observed is 0.1, moderate if it is observed as 0.3, and large if it is observed as 0.5. Criterion validity is validating with gold standard questionnaires or experts from that field to ensure the test measures the intended objective. It is also called concrete validity and measures the outcome of a questionnaire-based survey in different situations such as past, present, and future. Convergent validity tests the questionnaire for correlation with the previously validated tool that had been constructed for the same objective.
The magnitude of the sample size is critical for the research. Larger samples will waste the resources of the researcher and the organization, and smaller samples may not correctly represent the population under study. The sample size is computed by using a sample size calculator or use of qualitative guidelines like 5:1 (50 samples for 10 questions), 10:1, 15:1, or 30:1. A minimum of 300 patients should be assessed, and the value of 0.5 obtained from factorial analysis to validate the questionnaire the question matches an appropriate domain. A value of <0.5 requires repetition of the reliability test by framing new questions. The researcher may either discard the question or get an expert opinion on the necessity of retaining the question.
A researcher cannot use any question available online, as it may lead to copyright issues. The researcher should seek permission from the primary author to use it in their research without any commercial motive.
1. Aithal A, Aithal PS. Development and validation of survey questionnaire & experimental data – A systematical review-based statistical approach Int J Manage Technol Soc Sci (IJMTS). 2020;5:233–51
2. Onwuegbuzie AJ, Dickinson WB, Leech NL, Zoran AG. A qualitative framework for collecting and analyzing data in focus group research Int J Qual Methods. 2009;8:1–21
3. Aithal PS, Shailashree V, Kumar PM. Factors & elemental analysis of six thinking hats technique using ABCD framework Int J Adv Trends Eng Technol (IJATET). 2016;1:85–95
4. Taherdoost H. Validity and reliability of the research instrument; how to test the validation of a questionnaire/survey in a research Int J Acad Res Manage (IJARM). 2016;5:28–36
5. Lawshe CH. A quantitative approach to content validity Pers Psychol. 1975;28:563–75
6. Ayre C, Scally AJ. Critical values for Lawshe's content validity ratio: Revisiting the original methods of calculation Meas Eval Couns Dev. 2014;47:79–86