going after the data data collection instruments fetp india
TRANSCRIPT
The data collection instrument is a logical deduction of the analysis plan
Research question:? Risk factors for
leptospirosis
Study objectives:Estimate association
between water exposure and
disease
Design/ indicator:Case controlOdds ratio
Analysis plan:Dummy table
Data elementsNeeded:
? Water exposure? Sick
Data collection:Interview
Individual items:? Swam in water
? Sick
Consolidationof the
instrument
Information that may be collected with a data collection instrument
• Facts• Judgements• Indicators of knowledge
Instruments
Information that may be collected with a data collection instrument
• Facts Individual characteristics
• Height, age, income
Environment • Housing, family size
Behaviours, practices• Alcohol or tobacco consumption
• Judgements• Indicators of knowledge
Instruments
Information that may be collected with a data collection instrument
• Facts• Judgements
Opinions Attitudes
• Indicators of knowledge
Instruments
Information that may be collected with a data collection instrument
• Facts• Judgements• Indicators of knowledge
Risk factors Elements of healthy lifestyle
Instruments
Classical way to explore behaviours and their determinants in
epidemiology• Knowledge • Attitude• Practices
Instruments
Different ways to collect data with an instrument
• Abstraction form Clinical records Surveillance records Registers
• Structured observation guide• Questionnaire
Instruments
Triangulation to reconstitute the best possible reflection of the truth
• Collection of information on the same topic through various mechanisms
• Attempt to reconstitute a reliable reflection of the parameter
Instruments
Examples of triangulation to estimate the proportion of blood
units screened for HIV
• Interview of the laboratory manager Questionnaire
• ? What is the number of units screened
• Observation of the practices of the laboratory technician Structured observation guide
• ? Proportion of units tested
• Review of registers Abstraction form
• ? Number of tests ordered, used
Instruments
The four components of a data collection instrument
• Introduction and conclusion• Identifiers • Instructions for the person who collects
data• Body of the instrument
Items
The four components of a data collection instrument
• Introduction and conclusion Introduction
• Presentation, objectives• Elements needed for
informed consent Conclusion
• Identifiers • Instructions for the person who collects
data• Body of the instrument
Items
The four components of a data collection instrument
• Introduction and conclusion• Identifiers
Exact identifiers (e.g., name, address)• Collect and keep apart • Not entered in the computer
Coded ID number (composite)• Entered in the computer
• Instructions for the person who collects data
• Body of the instrument
Cluster
House
Person
Items
The four components of a data collection instrument
• Introduction and conclusion• Identifiers • Instructions for the person who collects
data Guide for the person who collects data Instructions (e.g., prompts) Skip patterns Use different fonts (e.g., italics)
• Body of the instrument
Items
The four components of a data collection instrument
• Introduction and conclusion• Identifiers • Instructions for the person who collects
data• Body of the instrument
Open items Closed items Semi-open items
Items
Different types of items in the body of a questionnaire
• Open questions The interviewer leaves the answer free
• Closed questions The interviewer proposes options of answers
• Semi-open questions The interviewer proposes options of
answers, but additional free answers are possible
Items
Open questions
• Answers are not suggested • Subjects must generate an answer• Advantages
Give freedom of response Stimulate memory Can be useful to generate closed responses later Useful at a hypothesis raising stage
• Inconvenient Difficult to code and analyze May be incomplete and / or unfocused
Items
Examples of open questions
• What disease can you acquire from tobacco?
• What places did you eat at in the week preceding the disease?
Items
Open questions with close ended answers
• No option of answer is suggested• However, among the answers freely
mentioned, the interviewer will tick those spontaneously specified
• Expressed as an open question • Analyzed as a close-ended question
Items
Example of open question with close ended answers
• What are the practices that may increase your risk to get a heart attack? (DO NOT propose any option of answer) Lack of exercise (Yes/No) Smoking (Yes/No) Poor dietary practices (Yes/No) Eating too much salt (Yes/No)
Items
Closed questions:1. Dichotomous options
• Suggested answers include “Yes” and “no”
• Advantages Forces a clear position May be useful for key, important, well
framed issues
• Inconvenient May oversimplifies issues
Items
Good and bad examples of closed dichotomous questions
• Have you ever consumed tobacco products? A dichotomous question here is likely to
over-simplify, unless it is used as an introduction
• Did you eat at restaurant X between 1 and 28 February?Adapted to an outbreak investigation
Items
Closed questions:2. Multiple options
• Multiple options of answers are suggested
• Advantage Larger choice of answer options
• Inconvenient May be difficult to choose only one option
Items
Examples of closed questions with multiple options
• Where do you go to seek treatment when moderately sick? (e.g., for fever) Hospital Public clinic Private clinic Pharmacist
• Do you wear a helmet when riding a bike? Always Sometimes Never
Items
Differentiating questions with multiple options from multiple
dichotomous questions • If more than one option of response, be
clear as to whether one or multiple answers are acceptable
• Only one answer acceptable=One variable with multiple options
• More than one answer acceptable=Equivalent to multiple dichotomous
variables
Items
Example of question with multiple options that lead to ambiguities
• What are the elements that led you to stop smoking? ? Fear of the danger of tobacco? Diagnosis of a tobacco related illness? Fear of dependence? Cost of tobacco products
• Two possibilities: Accept only one answer Accept multiple answers
Items
Possibility 1: More than one option acceptable
• What are the elements that led you to stop smoking? Fear of the danger of tobacco Diagnosis of a tobacco related illness Fear of dependenceCost of tobacco products
=Equivalent to multiple dichotomous questions, each option being a variable
Items
Clarified possibility 1: More than one option acceptable
• Among these elements, what are those that led you to stop smoking? Fear of the danger of tobacco
• Yes / No Diagnosis of a tobacco related illness
• Yes / No Fear of dependence
• Yes / No Cost of tobacco product
• Yes / No
Items
Possibility 2: Only than one option acceptable
• What are the elements that led you to stop smoking? Fear of the danger of tobacco Diagnosis of a tobacco related illness Fear of dependence Cost of tobacco products
=Equivalent to one question with multiple options of answers, one variable
Items
Clarified possibility 2: Only than one option acceptable
• Among these elements, what is the one that was most important in your decision to stop smoking? Fear of the danger of tobacco Diagnosis of a tobacco related illness Fear of dependence Cost of tobacco products
Items
Closed questions:3. Quantitative answers
• The subject must provide a quantified answer
• Advantage Allows creation of continuous variables
• Inconvenient May requires validation:
• Some “quantified” answers might be limited in the way they can be handled as continuous variables
Items
Example of closed questions with quantitative answers
• How many time did you visit the clinic in the last 12 months? True continuous variable Four visits is the double of two visits
• How would you describe your pain on a 1-10 scale where 1 would be the minimum and 10 would be the maximum? In fact a qualitative variable with 10 options Requires validation
• Six may not be the double of three on the scale
Items
Semi-open questions
• Suggested answers • Possibility to create another answer
Other, specify: __________
• Advantage Leaves the door open to unplanned answers
• Inconvenient Difficult to analyze
Items
Examples of semi-open questions
• Did you child have complication following measles? None Pneumonia Diarrhoea Eye problems Other, specify: ______________
Items
Formulating questions (1/2)
• Write short and precise questions Avoid ambiguities
• Use simple words of every day language• Avoid negations and double negations
Do you sometimes care for patients without washing hands?
Do you systematically wash hands before caring for each patient?
Production of the instrument
Formulating questions (2/2)
• Ask only one question at the time Did you refuse treatment because you feared side
effects? Did you refuse treatment? If yes, was this because you feared side effects?
• Be specific Are you aware of the modes of transmission of HIV? Among these practices, can you tell me those that
could lead to HIV?
• Use neutral tone to avoid influence Have you been promiscuous in the last six months? How many partners have you had in the last six
months?Production of the instrument
Sorting questions
• From the general to the specific • From the simple to the complicated• From the casual to the intimate• Regroup identification questions at the
beginning or at the end• Introduce simple questions as a break if
the questionnaire is complex• Triangulate through multiple questions
on the same topic if the subject is important
Production of the instrument
Careful lay out the data collection instrument: Rationale
• Easier to use• Guides the field worker• Reduces the risk of errors• Reduces the risk of forgotten questions• Simplifies coding • Simplifies data entry
Production of the instrument
Careful laying out the data collection instrument: Principles
• Split the sections• Space out questions• Use larger fonts• Align answers on the right hand side• Do not split questions across pages• Number questions• Standardize coding • Use auto-coding procedures
Production of the instrument
Auto-coding
• Q.25: Where did you go when your child had diarrhoea? 1. Hospital2. Public clinic3. Private clinic4. Pharmacist
2
Production of the instrument
Checking the instrument against the analysis plan
• Suppress unnecessary questions Those that do not be used in the analysis
• Add missing questions Those that will provide variables needed in
the analysis
Production of the instrument
Colleagues who can help in reviewing the questionnaire
• Colleagues• Experts• Statisticians (Coding)• Field workers • Data entry clerks
Production of the instrument
Language
• All questionnaires must be written in the language in which they will be administered Not acceptable to have an English questionnaire
translated in the field by the interviewers• No standardization
• Translation is required, with quality assurance Initial formulation (e.g., in English) Translation (e.g., in Hindi) Back-translation (e.g., back to English)
Production of the instrument
Objectives of the pilot testing of the questionnaire
• Check that the questionnaire is: Clear Understandable Acceptable
• Check flow and skip pattern• Check pertinence of coding • Estimate the time needed to ask all the
questions
Production of the instrument
Pilot testing the questionnaire in practice
• Pilot test with yourself • Pilot test with a few volunteers • Pilot test in real size
Persons similar to the study population Persons who are not to be included in the
study
Production of the instrument
Producing the last version of the questionnaire
• Professional finish• Paper of good quality• Interviewer’s kit
Sleeves Clip board Pencil, eraser
Production of the instrument
Summary of the systematic process leading to the data collection
instrument
Research question
Study objectives
Design/Indicators
Analysis plan
Data elementsneeded
Choice of data
collectionmethod
Formulation of individual
items
Consolidationof the
instrument
DANGER: By pass leads to poor studies
Take home messages
• Think instruments, not only questionnaire
• Prepare your items as future variables• Polish, polish and polish to ensure good
data quality