05/11/25 1

Data Collection
Tools

Dr. Muhammad
Shahbaz
GC Women University
Sialkot
05/11/25 2

Outline of this presentation

• Describe primary and secondary data
• Briefly describe the different data
collection tools
• Discuss the development of different
data collection tools in detail
• Describe the types of error that may
occur in data collection
• Discuss how to ensure validity and
reliability of data collection tools
Study Instrument
05/11/25 4
Types of research questions
we may take up
Research question Study design
• Describe the existing • Cross sectional
situation within the study
health system with a
view to improve
• Determine risk factors • Analytical study
for a health condition • Case control

• Assess the effectiveness • Cohort

of a procedure/ therapy
• Experimental study
• Adherence to prescribed
guidelines in patient
management • Clinical audit
Variables
• Something that varies.

• Factors / characteristics that are measured

in a research study

• E.g. Age, Sex, Weight, POA, Knowledge,

Work stress, Presence of a disease
05/11/25 6

Activity

• Write the main variables that need to be measured

in order to achieve your research objectives.
 Variables
• Classified in different ways:

• According to,
- Scale of measurement –
Nominal, Ordinal, Interval, Ratio scales

- Simple and Composite variables

- Objective and Subjective measurements

Nominal Scale
• Simply categorizes data into two or more
categories
• There is no mathematical relationship
between the categories

• E.g. Gender – Male / Female

Ethnicity – Sinhala / Tamil / Muslim
Ordinal Scale
• Categorizes data into two or more groups
which have an order or hierarchy

• Does not indicate how much larger one

group is from the other

• E.g. Social Class – Upper / Middle / Lower

Pain – Mild / Moderate / Severe
Interval Scale
• Measurements are assigned a value in a
numerical scale. The interval between
successive points are equal.

• Zero is arbitrary

• E.g. Temperature scale

Ratio Scale
• Interval between successive points are
equal.
• Has an absolute zero.

• Therefore the measurements can be

compared using ratios.

• E.g. Weight, Height. Hb level

Simple and Composite
Variables
• Simple variables measure obvious
characteristics that can be measured directly

E.g. Weight, BP

• Composite variables are complex variables

made up of a combination of simple variables.
• Often they are latent variables which cannot be
measured directly.

E.g. BMI, Stress level, Knowledge, Attitudes

Objective and Subjective
Measurements
• Objective measurements are directly
measured and well-defined.

E.g. Age, Parity, POA, weight, BP

• Subjective measurements are often

opinion-based and indirectly measured
concepts

E.g. -Attitude towards family planning

-Pain following LSCS
-Job satisfaction
Continuous and Discreet variables
• Applies for variables measure in interval
scale or ratio scale.

• Continuous variables can take any value in

between two points of the scale
E.g. Weight

• Discreet variables can take only discreet

values.
E.g. Parity, No. of births, No. of cases
05/11/25 15

Operationalizing variables
• Defining how the variables will be practically
measured for the purpose of this research study.

E.g. Dengue fever –

• Fever >2 days, Headache, Body aches,
Thrombocytopaenia
• Dengue Ig M Ab +ve
• Dengue Ag +ve

Knowledge on Epilepsy – Marks scored for 10 MCQ

questions on the disease, its treatment and
consequences
05/11/25 16

Activity

• Operationalize the main variables in your list

05/11/25 17

Data
05/11/25 18

Two types of data

• To be generated by the researcher

for the purpose of the study
Primary data

• Already available, generated by

someone else for other purposes
Secondary data
05/11/25 19

Secondary data
Primary data
Advantages Advantages
Disadvantages Disadvantages
05/11/25 20

Secondary data
Primary data
Advantages Advantages
• inexpensive/ data • can collect the whole
already available range of information
• permits examination of needed
information in the past • can collect precise /valid
data

Disadvantages Disadvantages
• information may be
• time consuming
imprecise or incomplete
• not easily accessible • resource intensive
• issues concerning
confidentiality
05/11/25 21

Secondary data
Primary data
Data extraction Questionnaires
form • Self administered
• Interviewer
administered
Checklists
• Observation
• Examination
Measuring
Equipment
05/11/25 22

Data collection tools

• Questionnaires
• Self-administered
• Interviewer administered

• Observation/Examination checklists

• Measuring equipment

• Data extraction sheets

05/11/25 23

Tool- Questionnaire
When to use?
presence/absence of an attribute
knowledge, feelings, opinions
attitudes

Two types
• interviewer administered
• self administered
 05/11/25 24

Interviewer administered Self administered

trained interviewers ask written questions are
presented to the respondent
questions and record to be answered in written
most commonly used in form
hospital/ clinic surveys/
community surveys used when
• a study population can read and
used when write and with similar educational
background
• participants have variable school children
ability to read and university students
understand questions professional groups
• if questions require • to collect information on sensitive
issues
explanations and guidance
05/11/25 25

Interviewer administered Self administered

Advantages Advantages
• can cover a large number
• suitable for a study
of respondents within a
population with varying short duration
educational • permits anonymity
background resulting in more honest
• suitable with illiterates responses
• not require assistants-
• permits clarifications
less expensive
• higher response rate • not influenced by the
presence of an interviewer
• can be even mailed
05/11/25 26

Interviewer administered Self administered

Disadvantages Disadvantages
• presence of interviewer • not to illiterate study
may influence-reporting populations
may not be accurate • low rate of response
• need trained interviewers • no way of clarifying
doubts
05/11/25 27

Tool- Checklist
Observation checklist- observe behaviour in
the natural setting and record

When to use?
assessing a skill, practice,
procedure

Researcher has to have a good knowledge

on steps of the procedure to design the
checklist
05/11/25 28

Observation checklist on hand

washing techniques
Steps description Completed
as described
Yes no
Step Wet hands and apply soap.
1
Step Rub palms together until soap is
2 bubbly.
Step Rub each palm over the back of the
3 other hand.
Step Rub between your fingers on each
4 hand.
Step Rub your hands with the fingers
5 together.
Step Rub around each of your thumbs.
6
05/11/25 29

• Examination checklist

presence of a clinical
sign/signs

eg: Joint
involvement in
arthritis
examine for inflamed
joints and mark
Tool- Measuring
05/11/25 30

equipment
Any measurement you do
• Anthropometry

• Clinical data
• Blood pressure

• Laboratory equipment /other equipment

• Haemoglobinometer
• Dip sticks
• Ultra sound scanner
Tool- Data extraction
05/11/25 31

sheet
A spread sheet/form to record the information
transferred from the secondary data source
Pt. age sex admissio reason diagnos
number n date for is on
admissi dischar
on ge
1
2
3
4
5
6
7
8
05/11/25 32
Any study design - one or
more of tools
Study design tools
• Cross sectional • Questionnaires
study • Self-administered
• Interviewer administered

• Analytical study
• Case control
• Observation/
examination checklist
• Cohort
• Measuring equipment
• Experimental study
• Data extraction sheets
• Clinical audit
05/11/25 33

Activity
• Identify a data collection instrument for
your study, giving reasons.
05/11/25 34

Errors
05/11/25 35

Types of errors that may occur in

research data
Sources of mistakes

•Interviewer/observer/person performing
the measurement/person performing the
laboratory test

•Respondent/person being observed or

measured

•Tools
 05/11/25 36
Sources of errors- Interviewer
Administered Questionnaires
Interviewer Respondent Questionnaire
•non-uniform reactions to the mistakes in
administration interviewer construction
• terms used  frightened
• probing  antagonistic/hostile
• assisting  suspicious
 annoyed
reactions to the
surrounding environment
fear of consequences/lack
of confidence of the use of
data
level of understanding
 05/11/25 37
Sources of errors- Self Administered
Questionnaires
NO Interviewer Respondent Questionnaire
•Eliminates this ability to read mistakes in
source of error  construction
level of understanding
reactions to the
surrounding environment
fear of consequences/lack of
confidence of the use of data
NO reactions to the
interviewer
 05/11/25 38
Sources of errors-
Observation/examination checklist
Observer/ Person being Checklist
examiner observed

mistakes in
•non-uniform
reactions to the construction
observation/ observer
examination  alter behavior
 05/11/25 39
Sources of errors- Measurements using
equipment
Person Person being Equipment
performing the measured
measurement/
laboratory test
physiological reactions
Non standardized
to the situation
faults
Non-adherence to diurnal variation
correct technique psychological reactions
to the interviewer
 frightened
 Excited
05/11/25 40

Improving validity of data

Measures against all In the stages of

sources of errors
•Deciding the tool/tools
•Interviewer/observer/
appropriate to obtain
person performing the
measurement/person information
performing the laboratory
test •Constructing the tools
and finalizing the tools
•Respondent/person being
observed or measured •Administering the
tool/doing the
•Tools measurement
05/11/25 41

Construction of
Questionnaires
• Content is transformed into questions and
instructions to the interviewer/respondent

• Decide on how the variables should be

operationalized to obtain the information
required
05/11/25 42

Construction of
questionnaires
Questions are basically two types

• Open-ended

• Closed-ended
05/11/25 43

Open-ended questions
‘What are the problems you experienced when
taking
drug X ?’ ……………………………………………

When to use?

• if the information required is a numeral

• when it is important to have information on how

the response was expressed
• has limited knowledge on what the responses
would be
• when checking for ‘core’ knowledge
05/11/25 44

Closed-ended questions
‘Which of the following problems did you
experience
when taking drug X?’
Drowsiness
Vomiting
Itching

• Choose from a number of fixed

alternatives which should be
• Exhaustive/comprehensive
• Mutually exclusive
05/11/25 45

Closed-ended questions
Open-ended questions

Disadvantage Disadvantages
• not allow own answers
• may generate a wide which would be potentially
range of responses more accurate
• not allow detailed responses
• difficult to categorize
in analysis
Advantages
• quicker and easier to answer
• responses are already
Advantages categorized for analysis
• Can get detailed • list of answers help clarify
responses the question
05/11/25 46

Mixed type
Choose from a number of fixed alternatives
and an added opportunity to account for
missed choices

‘Which of the following problems did you

experience when taking drug X?’
Drowsiness
Vomiting
Itching
Any other (specify)
………………………
05/11/25 47

Multiple responses
Single response
‘What was the problem that you • Cannot expect by the
experienced mostly when taking
drug X?’ respondent unless
you instruct

• Which of the
following problems
did you experience
• When have to select one when taking drug X?’
response from a range
• read all
ye
• show on a card s no
Drowsiness
Vomiting
Itching
05/11/25 48

Visual analogue scales

• Mark a spot that best represent his characteristic

on a line representing a continuum from one
extreme to the other
• anchor each end with words that describe the
most extreme values

‘Please mark an X on the line, the place that best

describes the severity of your pain during the first
day after surgery’

no pain at all
unbearable pain
05/11/25 49

Likert scale
• Used in instances in which responses
represent a continuum

• E.g.: Assessing satisfaction

Assessing attitudes
05/11/25 50

Statements Strongl Agree No Disagree Strongl

y comm y
agree ent disagre
e
Fat kids are healthier
than thin ones
one cannot do much
about obesity even if
tried hard
• Respondents attitudes may be latent/ never
have given a thought till he was confronted
with the question
• Requires thought and self analysis
• Many sides to the issues assessed
• moral, medical, legal
• answer depend on aspect that is uppermost in his
mind
05/11/25 51

Other principles
• Avoid acronyms, abbreviations and
technical terms, use simple, common
words

eg. ‘Sputum vs phlegm’

Contraception vs family planning
05/11/25 52

Other principles
• Avoid
Leading questions
‘Do you agree that exclusive breast feeding is
good for your baby?’

Questions with hidden assumptions

‘How many cups of milk do you drink a day?’

Double negative questions

‘didn’t you breast feed the baby on the days that
the baby was not having fever?’
 05/11/25 53

Other principles
• Avoid
Unanswerable questions
Do you approve or disapprove the new law to legalize
abortion ?
-Yes
-No

Double Barrel questions

Are you satisfied with the salary and social status you
receive from this job ?
05/11/25 54

Other principles
Maintain neutrality

•Set a tone that permits respondents to

admit to behaviours and attitudes that may
be undesirable

‘People sometimes forget to take

medication as instructed. Does that
ever happen to you?’
05/11/25 55

Other principles

•Give ‘don’t know’ and ‘not applicable’ options

•Use adequate words.

Age?
Age in completed years.

When did you have measles?

At what age did you have measles?
05/11/25 56

Other principles
Set time frame
• If designed to measure the
• frequency of an event
‘How many episodes of loose stools did your
child experience?’

• frequency of certain habitual or recurrent

behaviour
• On average….how many times or on a usual day
how many times…..

• Questions needing recall- also offer a ‘cannot

remember’ option
05/11/25 57

Other principles
When assessing knowledge,

• Open questions vs Closed questions

• In closed questions offer a ‘don’t know/
cannot remember’ option
• When using MCQ type questions – Make sure
there is only one correct answer
• Scoring of Knowledge questions – Equal
scores vs Unequal scores
 05/11/25 58

Other principles
When assessing attitudes

Allow adequate options to answer.

E.g. Totally agree Agree Unsure Disagree Totally
disagree

Keep the statements simple

E.g. With the economic conditions as the way they are these
days, it really is not fair to have more than two children in a
family.

Avoid leading questions

Are you against giving equal rights to males and females?
 05/11/25 59

Other principles
When assessing attitudes
Avoid double scales
E.g. How would you rate your job ?
- Rewarding but stressful - Rewarding and stress-free
- Unrewarding but stress-free - Unrewarding and stressful

Use non-judgmental statements.

Do you believe poor people should not be given free health
care ?

Do you believe there should be a universal fee-levying system in

our health care service ?
Reduces Social Desirability Bias
05/11/25 60

Other principles
• Clear instructions

• Self administered questionnaires to the

respondent
• how to fill

• Interviewer administered questionnaire to the

interviewer
• To be read out to the respondent
• To remind important instructions
05/11/25 61

Other principles
• Clear instructions to respondent/ interviewer
• how to mark the responses
• whether more than one response allowed
• to skip irrelevant questions

yes no
10. Have you ever been
told that you have
diabetes’
if ‘No’ go to q. …

11. At what age were you diagnosed? …………

• have section breaks

 05/11/25 62

Language of the questionnaire

• Should be understood by the respondent
• Should be in the language of the interview
• Simple/ everyday language

Format of the questionnaire

• layout
• spacing
• font size
• not too long
05/11/25 63

Constructing other tools

• Observation checklists/Examination
checklists
• include all steps/aspects
• take care to use terms/ language to suit the
observer/examiner
• instructions for the observer/examiner

• Equipment
• select most refined
• Standardize prior to and in between
• as far as possible automated
05/11/25 64

Validity
Appraising validity of the draft
tools
Any tool
• Face validity/Content validity- subjective
judgment whether it covers all steps

• Consensual validity- whether experts in the field

agree that the tool covers all aspects of the
phenomena under interest

• Pre-testing of tools- modify and clarify,

add/drop items, timing

• Reliability- better to report

05/11/25 65

Appraising validity of the

draft tools
• May need a validation study if the tool is
• assessing a concept which cannot be directly
assessed
• eg: parenting stress
psychological distress
Quality of life

• developed in another country EVEN if the

tool is validated in that country
• translate to local languages
• adapt to suit the culture
05/11/25 66

Internationally validated
tools
• validity depend on
• language
• culture
• socio-demographic characteristics of the
respondent
• mode of administration
• etc….
05/11/25 67

Reliability
• Test-retest • By applying the tool
reliability after a time gap
• Select a few
important variable
or two

• Internal
consistency • If it is a multi-item
• Cronbach’s alpha scale
Measures taken to minimize
05/11/25 68

errors due to interviewers/

observers
• Minimum required number
• Training
• Supervision
• Blinding
 Measures taken to minimize
05/11/25 69

errors due to interviewers/

observers
• Minimum required number of
interviewers/observers/technicians with equal
qualifications
Measures taken to minimize
05/11/25 70

errors due to interviewers/

observers
• Training- interviewers, observers, technicians
• interactive sessions, role play
• uniform administration of questionnaires
• terms used
• probing
• assisting
• uniform observation/ examination
• uniform and correct technique in using
equipment

specify steps to follow /instruction manual

Measures taken to minimize
05/11/25 71

errors due to interviewers/

observers
• supervision

• blinding- making observers blind to the

research question
Measures taken to minimize
05/11/25 72

errors due to respondent/

person being observed
 introductions and explanations of the study-
informed consent, confidentiality
 neutral environment

 take measures to assure anonymity

 use familiar/known person as the interviewer

/examiner
 physical preparation of the patient
05/11/25 73

Thank You