Data collection methods & tools

George Mwasekaga, MD

1 1/17/2024
 Presentation outline
 Learning objectives
 Introduction
 Data collection methods
 Data collection tools
 Questionnaire

2 1/17/2024
 Learning objectives
 Identify types of data collection methods
 Identify advantage and disadvantage of each data collection
methods
 Describe step by step data collection process
 Identify data collection tools
 Develop a Questionnaire

3 1/17/2024
 Introduction
 Having decided on your study design
 Next methodological decision is how to collect information
 Most commonly methods for data collection;
Interviews
Questionnaires
Observation

4 1/17/2024
 Interviews
 Usually held on one to one basis, group interviews or focus
groups
 Can be highly structured, semi-structured or unstructured
 The degree of structure affects the flexibility of the interview

5 1/17/2024
 Questionnaires
 Comprised a written questions that are answered by all the
respondents in a study
 Several different types questions can be used
Closed questions seek a limited response
Open questions allow respondents to answer freely on
their own words
 Open questions used when a more extensive response is
needed. Example: an explanation
Questionnaires are used to asses attitudes

6 1/17/2024
 Observation
 Technique for collecting data through visual observation of
events
 It requires the nature of the data to be observable
 Observation schedule can be highly structured or relative
unstructured
 Depending on information required and nature of observed
event
 Note: The method of data collection chosen for the study
should be appropriate for the type of information required

7 1/17/2024
Interview
Advantages Disadvantages
 No items are overlooked  Takes time to arrange
 Questions and answers can  Time consuming as it takes
be clarified by both longer time to collect data
interviewer and
interviewee  Traveling can be costly
 Researcher can achieve
depth of response
 Can be used to probe
sensitive or difficult areas
 Good response rate

8 1/17/2024
 Interview cont’d
 Interviewees don’t need to be  Interview skill are needed
able to read or write  Risk of interview bias
 Interviewees responses are not  Data analysis can be time
influenced by reading ahead (as consuming
with questionnaire)
 Responses are enriched by
observing non-verbal response
 Can be used as an exploratory
stage in large study

9 1/17/2024
 Questionnaire
Advantages Disadvantages
 Simple method to collect  Can not probe a topic in
data depth
 Rapid and efficient method  Respondent can omit items
of gathering data without explanation
 Collect data from widely (incomplete data)
scattered sample  Selection of forced choice
 Collect data from large item may insufficient to
sample reflect respondent choice

10 1/17/2024
 Questionnaires cont’d
Advantages Disadvantages
 Relative inexpensive  Amount of information
 Respondents can remain limited by respondents
anonymous interest and attention
 Easiest tool to test  Questionnaires can go
reliability and validity astray
 Respondent has time to  Production and distribution
consider each question can be expensive
 Sample is limited to those
with literate skills

11 1/17/2024
 Questionnaires cont’d
Advantages Disadvantages
 Analysis of data can be done  Most people express
quickly themselves better through
the spoken word
 Can be used to collect data
 No opportunity for
on a wide range of
researcher to interact with
topics/attributes respondents
 If respondents are
anonymous they can not be
followed up

12 1/17/2024
 Observation
Advantages Disadvantages
 Best way to record human  Time and duration of an
behaviour event may not be
 Observations recorded as predictable; involve
they occur eliminating watching and waiting
recall bias  Presence of observer adds a
 Allow research to view a new dimension to the
situation in total and in situation
context  Presence of observer can
affect peoples behaviour

13 1/17/2024
 Observation
Advantages Disadvantages
 Observation may be subject
 Observation schedules can
to observer bias
be simple to design and use
 Observers may find
 Observers may need little
themselves drawn into the
training
situation
 Open to the use of
 Event may occur so rapidly
recording devices
it not to record everything
 Little control over number
of times an event will occur
 Those not wishing to be
involved may object to the
presence of the observer
14 1/17/2024
 Important note on questionnaire
 It is not necessary to design a new data collection tool for a research
project
 Researcher may discover a questionnaire that suit the intended
purpose during literature search
 It is acceptable to use data collection tools that have been developed
by previous researchers
 This may be preferable especially if the tool have been subjected to
test for reliability and validity

15 1/17/2024
 Questionnaire design
 Collection of relevant data are an essential part of any
epidemiological study
 Data will include identification information such as:
Name, Age, Sex, Place of residence
 Information on the main outcome and exposures
Types of data we can collect
 Clinical data and lab data often linked to: Socio-demographic
data
 Behavioural data
Often different sources of data are used

16 1/17/2024
 COMPOSING QUESTIONS
 Good questions are hard to write and require:
1. A clear understanding of the key objectives of the study
2. Skills and experience or expert advise
3. Careful thought and practical testing
 The answers may vary widely, depending precisely how the
questions are constructed and worded
 “Open” and “closed” questions
Open – no pre-determined set of responses
Closed – like multiple-choice; usually 5-10 options

17 1/17/2024
  All respondents should understand questions in the same way
 Must be clear and unambiguous
 Yes/No/Don’t know answers
Ask for answer to one thing at a time
e.g. Do you take your child to a doctor when he has diarrhoea
or a cold?
 Don’t use technical jargon e.g.
When did you have your last myocardial infarction?
Better…When did you have your last heart attack?
 Avoid leading questions e.g. “What are the main thing that is
wrong with the care that you get from your doctor?”
Could ask “Do you like the care you get from your
doctor?” (Y/N/No opinion) or What do you think of the care
18 you get from your Dr? (Excellent, poor etc) 1/17/2024
 SENSITIVE QUESTIONS
 It may not be possible to avoid asking questions that may
embarrass or offend some respondents
 Are your questions acceptable?
 Is it practical to expect to be able to get this information?
 Put sensitive questions towards the end of the questionnaire
 Indirect questions can be used to obtain information on
sensitive issues
 May be appropriate for data on income, sexual norms or
socio-economic status

19 1/17/2024
  Broad category of response may be used if this satisfies the
needs of study
e.g. “What do you think is the average amount
people earn per month in this village?”
 Possible offense / embarrassment can prevented by including
a statement designed to show interviewer’s interest is non
judgmental e.g. “ We know that married couples sometime
quarrel with each other. How often do you quarrel with your
husband?” e.g. in a school- boy interview ‘As you know, many
have cheated in their exams during their life. Did you ever
cheat in an exam?’ or ‘Do you know students who cheated in
their exams last summer?’[ b ] what about yourself?

20 1/17/2024
  Section 4: Sexual Practice
Read: Now I would like to ask you some questions about sexual
activity in order to gain a better understanding of issues related
to reproductive health. I know these questions may be sensitive,
but I hope you will respond to help us to better understand
some health and childbearing issues.

21 1/17/2024
 PREPARING QUESTIONNAIRES
 Back-translation of drafts
Translate questions into the local language and back translate
into original language e.g. English to Swahili and back to
English
 Very important to check context of questions and content
have been preserved

22 1/17/2024
 PILOT TESTING
 The translation of the questions may be validated in the pilot
 In some studies a pilot studyof the questionnaires is
incorporated into the training process for the interviewers
 Enables supervisors to assess the quality of the filled
questionnaires across the interviewers

23 1/17/2024
 Quality control
 Quality control is needed in all aspect of a study:
For questionnaire response
For laboratory testing
For clinical examination
 Some studies select 10% of subjects to go through a quality
control exercise
 Staff should be aware that their performance is being
monitored
 Results of the quality control exercise needs to be conveyed
to the interviewers
 If performance is not good enough interviewers should be
retrained or replaced
Monitor interviewe rperformance early or it will be too late
24 1/17/2024
 Preparation for data collection
 Logistics
 Forming a research team
 Preparing research documents
 Training the research team
 Supervision and documenting the process

25 1/17/2024
 Logistics
 Will participants be visited/examined at their homes or
asked to come to a central location?
 How will samples be stored and transported to a laboratory?
 How will people be traced?
 How will confidentiality be maintained?

26 1/17/2024
 Research team
 What man power is necessary to do carry out the research?
 Think through the logistics of the desired field team:
How will they flow in the research center?
Can they be accommodated in the vehicle along with
equipment?
 Develop job descriptions for personnel including educational
requirements and experience
 Organograms- helpful to show relationships and lines of
authority
 Recruit research staff

27 1/17/2024
 Field documents
 Protocol
 Manual of Operations
(MOP)/ Field Procedure
Manual
 Standard Operating
Procedures (SOPs)
 Questionnaire
 Tracking Logs

28 1/17/2024
 Tracing log
 Screening & Enrollment
 Visit Logs/Appointment
calendars
 Drug dispensing
 Lab chain of custody

29 1/17/2024
 Training of research team
 Essential for producing quality
research results!
 The more knowledge the team has on
the study, the more likely they will do
quality work.
 Don’t underestimate the role of any
member on the team. Everyone is
contributing toward producing quality
research.

30 1/17/2024
 Supervision & documentation progress
 Its important to develop systems of monitoring quality
assurance
 Assuring SOPs are followed
 Staff are complying with protocol, guidelines and
trainings
 Quality checks on data

31 1/17/2024
References
 Antony Stewart(2002). Basic Statistics and
Epidemiology A practical guide. Radcliffe Medical
Pres
 Bonita, R. Beaglehole, & T. KjellstrÖm (2006).
Basic epidemiology World Health Organization
 Bland, M. (2000). An introduction to medical
statistics (No. Ed. 3).Oxford University Press.
 Johnson, B., & Christensen, L. (2008).
Educational Research: Quantitative, qualitative,
and mixed approaches. Sage.