Dear Noel,
Namaskar.
I am happy that you are going to develop a fresh questionnaire. I hope that once it is finished, you will post it for others to use. Assuming you are familiar with psychometry and questionnaire construction, I am providing you with the steps below.
Step 1 - Collection of Content:
Go to your potential study population and approach about 50 individuals. Provide them with a plain paper and ask them to write 3 positive statements and 3 negative statements on the theme for your questionnaire. It is acceptable if some cannot provide all 6 statements. Collect all the papers.
Step 2 - Preparation of the Preliminary Questionnaire:
Convert the statements you have collected into question forms. Ensure that all questions are direct, easily understandable by the subjects, concise, unambiguous, and designed to elicit the desired answers. Typically, questionnaires include yes and no categories. Aim to have 100 to 150 questions in this preliminary questionnaire. The instructions for the subjects should precede the questions.
Step 3 - Collection of Data for Standardization:
Administer the questionnaire to a significant sample size and collect responses. While there is no specified standardization sample size, experience shows that statistical values stabilize when the sample size exceeds 200.
Step 4 - Item Variance:
Tabulate the frequency of responses in each response category for every item. Calculate the item variance by determining the probability of responses in each category (p and q for binary responses). The item variance ranges from 0 to 0.25 for two response categories. For more response categories, use quartile deviation to assess dispersion values.
Step 5 - First Selection of Items:
Decide on the number of questions to include in your final questionnaire. Begin by selecting questions starting from the highest value, choosing slightly more than needed as some may be eliminated in the next step. Steps 4 and 5 aim for a normal distribution of sample scores.
Step 6 - Tentative Scoring Direction and Internal Consistency:
Decide on the scoring direction for responses (e.g., higher value for yes, lower for no). Calculate a tentative score for each response and compute item-total correlation to assess consistency. Adjust scoring direction for items with negative correlations and repeat the process. Select questions based on correlation values until reaching the desired number for the final questionnaire.
For enhanced internal consistency, consider factor analysis.
Step 7 - Distribution and Norms:
Analyze the test score distribution to achieve a normal distribution. Establish percentile norms.
Steps 8 and 9 - Establish Reliability and Validity of the Questionnaire:
Ensure to establish the reliability and validity of the questionnaire.
Feel free to ask for clarification if needed.
Regards,
Jogeshwar