We have our plan (Chapter 14), and now we need the actual box to collect our information—the Data Collection Tool (like a super-questionnaire). If the box is badly built, the answers will fall out or get twisted, making our SuperPaste study useless. A good tool ensures the data we collect is valid and reliable.
1. What Information Are We Collecting?
We primarily collect three types of information in health research:
- Facts: Characteristics of the children, their environment, their behaviors (like how often they brush), and their practices (like using the paste).
- Knowledge: Their understanding of things like why Dental Caries happens or what a healthy lifestyle is.
- Judgments: Their opinions or attitudes (e.g., whether they like the taste of SuperPaste or their opinion on the quality of the local dental service).
2. The Tools We Use
Depending on what we are collecting, we choose the right tool:
| Tool | Simple Idea | When to Use It |
| Abstraction Forms | Culling information from existing paperwork. | Looking at clinical records, disease surveillance data, or school registers to get facts on past Dental Caries. |
| Structured Observation Guide | A checklist to confirm processes are done correctly. | To document if the children are actually using the prescribed brushing technique and for how long. |
| Questionnaire | Asking the participants directly. | Most common method, used to collect facts, knowledge, and judgments. |
- Interviewer Administered: Data collector asks questions (face-to-face or phone).
- Self-Administered: Participant fills it out themselves (only for literate and knowledgeable people).
3. The Structure of the Questionnaire
Every good data collection tool has four key parts:
| Component | Purpose | Ethical Consideration |
| Introduction | Presents the study, states objectives, and obtains Informed Consent. | Sets a professional tone and ensures participation is voluntary. |
| Body (Question Items) | The main content where all the data is collected. | Must use clear language, be specific, and avoid ambiguity. |
| Instructions | Directions for the interviewer (e.g., Skip Patterns like “If No, go to Q19”) and general prompts. | Helps the study flow smoothly and saves time. |
| Identifiers | Codes or names to track the participant. | Use coded ID numbers (e.g., composite codes for state-district-individual) instead of actual names to maintain confidentiality. |
4. The Types of Questions
The way we ask a question changes the kind of answer we get:
| Question Type | Simple Explanation | Advantage | Disadvantage |
| Open Questions | The answer is not suggested; the participant gives a free response. | Gives total freedom, good for generating hypotheses (unplanned answers). | Difficult to code and analyze later; responses may be unfocused. |
| Closed Questions (Dichotomous) | Only two options are provided (e.g., Yes/No, Male/Female). | Forces a clear, focused position; useful for key information. | Can oversimplify complex issues. |
| Closed Questions (Multiple Choice) | Multiple options are provided; could allow single or multiple answers. | Easy to code and analyze; ensures completeness. | Requires clear instructions on whether only one answer is acceptable. |
| Closed Questions (Quantitative) | The answer must be a number (e.g., “How many times did you brush last week?”). | Allows creation of continuous variables (measurable data). | Can be hard for participants to recall or provide an exact number. |
| Semi-Open Questions | A list of suggested answers is given, plus an “Other (Specify)” option. | Leaves the door open for unplanned answers without forcing analysis chaos. | If too many choose “Other,” analysis becomes difficult. |
5. Rules of Phrasing (Do’s and Don’ts)
| Do’s | Don’ts |
| Use short and precise questions (e.g., “What is your age?”). | Use complicated academic language; use everyday words. |
| Use a neutral tone and avoid judgmental words (e.g., “promiscuous”). | Use negatives, especially double negatives (“Do you sometimes care for patients without washing hands?”). |
| Ask only one question at a time (avoid asking about a fact and the reason in one sentence). | Ask vague questions (e.g., “Are you aware of the modes of transmission of HIV?”). Be specific! |
| Group together all questions related to the same topic. | Split questions across pages. |
| Use a chronological order when asking about a sequence of events. | Cramp questions together; use spaces and large fonts (11 or 12). |
6. Layout, Coding, and Final Checks
A. Layout and Coding
- Alignment: Align questions on the left, and answers/codes on the right (two-column look).
- Coding:Inbuilt coding system is essential for data entry.
- Standardize: Always use the same code (e.g., $1$ for Yes, $0$ for No).
- Auto-Coding: Use the number of the answer category as the code (e.g., if option 2 is chosen, the code is $2$).
B. Translation and Pilot Testing
- Translation: Translate the English version into the local language. Have someone else perform a back translation into English to ensure the meaning hasn’t changed.
- Review: Have colleagues, experts, and even the field workers review the tool for clarity and flow.
- Pilot Test:Crucially, administer the final tool to a few volunteers similar to the study population.
- Purpose: Check if questions are understandable, skip patterns work, coding works, and how much time the interview takes.
- Note: The subjects used for the pilot test must not be included in the main study.
Remember: The study questionnaire can make or break the study. Collect valid and reliable data the first time, as there is often no opportunity to go back!
Leave a Reply