When conducting research, the strength of the conclusions largely depends on the quality of the research instrument used for data collection. A poorly designed questionnaire can result in misleading or unusable findings, while a well-validated instrument ensures accuracy, credibility, and replicability. Among the key quality checks, validity plays a critical role.

Validity answers the question: Does the instrument actually measure what it is supposed to measure? For example, if a researcher designs a questionnaire to measure financial literacy, the items must reflect actual knowledge and decision-making in finance rather than unrelated topics such as general intelligence or personal preferences.

This blog explains the different types of validity, steps to establish them, and practical examples drawn from questionnaire-based studies across education, healthcare, psychology, business, and marketing.

Types of Validity and Their Application in Questionnaires

1. FACE VALIDITY

Face validity refers to how appropriate and relevant the questionnaire items appear to respondents or experts. While it is subjective, it provides a first-level check.

Example (Education Domain):

A researcher develops a questionnaire to assess students’ attitudes toward online learning. Items such as “I find online learning platforms easy to use” and “I can interact effectively with instructors online” clearly reflect the construct. However, if the questionnaire includes unrelated items like “I enjoy outdoor sports,” respondents may feel the tool lacks relevance. Thus, expert review ensures that items “look like” they belong.

2. CONTENT VALIDITY

Content validity ensures the questionnaire covers all the essential dimensions of the construct. This is usually achieved through expert judgment and measured using the Content Validity Index (CVI), where experts rate each item for relevance.

Example (Healthcare Domain):

Suppose a questionnaire is designed to assess patient satisfaction in hospitals. Experts identify five core domains: doctor’s communication, nursing care, cleanliness, waiting time, and billing process. The researcher creates items for each dimension and asks healthcare professionals to review coverage. If “billing process” is omitted, content validity becomes questionable.

Table 1: Expert Evaluation for Content Validity Index (CVI)

Item Statement	Expert 1	Expert 2	Expert 3	CVI (Proportion of Agreement)
Doctors explain treatment clearly	Relevant	Relevant	Relevant	1.00
Nurses are polite and approachable	Relevant	Relevant	Relevant	1.00
Hospital waiting time is reasonable	Relevant	Relevant	Not Relevant	0.67
The billing process is transparent	Relevant	Relevant	Relevant	1.00

A CVI of 0.80 or above is typically acceptable.

3. CONSTRUCT VALIDITY

Construct validity examines whether the questionnaire actually measures the theoretical construct. This is often tested statistically using Exploratory Factor Analysis (EFA) or Confirmatory Factor Analysis (CFA).

Example (Psychology Domain):

A researcher designs a questionnaire to measure workplace stress. Theory suggests three dimensions: workload, role conflict, and lack of support. Through EFA, items such as “I often have too much work to finish on time” load on the workload factor, while “I receive conflicting instructions from different supervisors” loads on the role conflict factor. If the factor structure matches theory, construct validity is confirmed.

Table 2: Factor Loadings from EFA

Item Statement	Workload	Role Conflict	Lack of Support
Too much work to finish on time	0.82	0.10	0.05
Conflicting instructions from supervisors	0.08	0.75	0.12
Lack of cooperation from colleagues	0.05	0.11	0.80

4. CRITERION-RELATED VALIDITY

Criterion-related validity shows how well a questionnaire correlates with an external criterion or “gold standard.” It can be concurrent (same time) or predictive (future outcomes). This validity is very useful and often applied in academic research, especially when researchers want to test how well a questionnaire, test, or scale predicts or correlates with an external measure (the criterion). This is particularly valuable when the research aims at prediction, evaluation, or decision-making.

Example (Business Domain):

A company develops a questionnaire to measure customer loyalty.

For concurrent validity, the scores are compared with customers’ actual purchase frequency from the past six months.

For predictive validity, the same scores are compared with purchase behaviour after six months.

If loyalty scores strongly correlate with real purchase data, the questionnaire demonstrates good criterion validity.

Table 3: Correlation with Purchase Behaviour

Customer Loyalty Score (Survey)	Purchases in 6 months (Criterion)	Correlation (r)
85	12
72	9
55	5
40	3	0.79 (Strong)

5. CONVERGENT AND DISCRIMINANT VALIDITY

These are finer aspects of construct validity, typically assessed through Confirmatory Factor Analysis (CFA) and Structural Equation Modelling (SEM).

• Convergent Validity: Items measuring the same dimension should correlate strongly.
• Discriminant Validity: Constructs that are theoretically different should not correlate too highly.

Example (Marketing Domain):

A questionnaire is developed to measure brand engagement and brand trust.

Convergent validity is established if items under “brand engagement” (e.g., “I actively follow this brand on social media”) are highly correlated.

Discriminant validity is confirmed if “brand engagement” is not overly correlated with “brand trust,” since they are related but distinct constructs.

Steps to Conduct Validity for Research Questionnaires

1. Define the Construct

Begin with a clear theoretical framework. For example, if measuring financial literacy, define whether it includes knowledge, behaviour, and attitude.

2. Generate Items

Draft questionnaire items based on literature review, expert input, and field understanding.

3. Check Face and Content Validity

Have experts review items for appropriateness and comprehensiveness. Use CVI to quantify judgments.

4. Pilot Test

Administer the questionnaire to a small group. Revise ambiguous or irrelevant items.

5. Statistical Testing (Construct Validity)

Use EFA/CFA to confirm whether items load onto expected factors. Assess reliability alongside validity.

6. Criterion-related Validation

Compare questionnaire scores with real-world outcomes or existing standardized instruments.

7. Report Findings

Clearly explain the process and results in the methodology section to strengthen research credibility.

Cross-domain Examples of Questionnaire Validity

To better understand how validity applies across disciplines, the following table provides general illustrations of how different types of validity can be applied when designing and testing research questionnaires.

Table 4: General Applications of Questionnaire Validity Across Domains

Domain	Type of Questionnaire	Validity Focus	General Application Example
Education	Student Learning Attitude Survey	Content Validity	Ensuring items cover all aspects such as motivation, interest, and classroom participation
Healthcare	Patient Satisfaction Questionnaire	Face & Content Validity	Experts confirm that items reflect key service dimensions like care quality, waiting time, and communication
Psychology	Stress or Self-esteem Scale	Construct Validity (EFA/CFA)	Factor analysis used to check whether items group under theoretical constructs (e.g., workload, self-worth)
Business	Employee Engagement Survey	Criterion-related Validity	Scores are compared with performance ratings or turnover data to confirm predictive power
Marketing	Customer Loyalty or Brand Engagement	Convergent & Discriminant Validity	Statistical tests confirm related items converge while unrelated constructs remain distinct

Which Validity is Most Important and Common in Research?

In academic research, all forms of validity are important, but some are more commonly adopted than others, depending on the research design and purpose. Here’s a structured comparison:

• Content validity is the first step to ensure conceptual coverage.
• Construct validity (with convergent & discriminant validity) is the most important and most commonly adopted in academic research, especially when studying abstract constructs like attitudes, perceptions, or behaviours.
• Criterion-related validity is important in predictive studies but less universally used.
• Face validity is supplementary and used more for practical reassurance but not strongly emphasized in academic publishing.

In practical terms, many researchers feel safer with content validity because it doesn’t require huge samples (200 to 300) or advanced statistics (CFA, SEM, etc.). However, the risk of waste of time and resources can be minimized with pilot testing, pre-validation, and iterative refinement in content validity.

Content validity is easier, reliable, and ensures conceptual soundness through expert judgment, while construct validity provides empirical proof but requires more resources. For academic rigor, both are essential—content as the foundation and construct as the confirmation.

Conclusion

Validating a questionnaire is essential to ensure that it accurately captures the construct under investigation. While face and content validity provide initial assurance, statistical methods such as EFA, CFA, and correlation tests establish deeper evidence of validity. Practical applications show that educational researchers, healthcare professionals, psychologists, business analysts, and marketers all rely on different forms of validity to refine their questionnaires. Content validity (foundational) and Construct validity (through convergent & discriminant validity checks) are the most important and usually adopted in academic research.

A validated questionnaire not only strengthens research findings but also enhances credibility and replicability across contexts. Therefore, validity should never be treated as an afterthought but as an integral part of research design.