Malimu intro to surveys

Definition of survey
 Study involving a subset (sample) of
individuals selected from a larger
population
 where variables are observed or measured
at a single point in time
 aggregated over all individuals in the
sample
 to obtain a summary statistic

Observational studies
exposure outcome
outcome exposure
outcome exposure
cohort
Cause-effect
case-control
Effect-cause
cross-sectional

Extension of cross-sectional design
 Longitudinal
 multiple cross-sectional surveys
 can estimate incidence
 if results between multiple surveys differ
 true difference?
 samples very different?
 Panel design
 multiple cross-sectional surveys using exactly
same sample
 Pseudo-cohort
 cohort: one age group over time
 pseudo-cohort: different age groups studied cross-
sectionally at one point in time

Survey
 To investigate
 if taxi drivers are more likely to use seat belts
than drivers of private automobiles
 To determine the community prevalence of
 malnutrition
 measles
 Tuberculosis
 HIV

Steps in designing a survey
Define survey objectives
Identify study population
Identify and define study variables
Establish plan of analysis
Determine sample size
Choose sampling method
Decide on mode of data collection
Design, field test and revise questionnaire
Train interviewers, conduct field work
Check all interview forms for error
Enter, tabulate and analyse the results

Survey objectives
 Formulate the objectives of the survey:
 What are you interested in finding out?
 Who do you want to study?
 When do you want to do the survey?
 What do you expect to learn and why?

Attributes of “SMART” objectives
S pecific
M easurable
A ction oriented
R ealistic
T ime frame

Example 1: Objectives
What? What are the knowledge, attitudes
and practices (KAP) to HIV/AIDS
Who? of adults > 18 years of age
Where? in Mwananyamala
When? in year 2006
Why? prior to the start of a new health
education programme

Example 2: Objectives
To measure and compare the prevalence of
smoking among MOH and Ministry of
Agriculture employees in Tabora
To measure and compare the proportion of
non-smokers in MOH and MOA exposed to
tobacco smoke in their working environment
and to determine the extent of the exposure

Define survey objectives 
Identify study population

(Target-) Population
Sample (study population)
Study population

Study population
 Objectives  target population
 Study population must be representative of
target population
 Time: seasonality, day of the week, time of the
day
 Place: urban, rural
 Persons: age, sex, other characteristics
 A non-representative sample will produce
biased results (validity )

Identify study population 

Study variables
 Identification
 as many as necessary but as few as possible
 each must “pay its own way”
 if of no use in analysis, do not include variable
 Requirements
 clear definition
 method of measuring should be:
 repeatable
 valid

Identify study variables
 Objective: To measure and compare the prevalence of
smoking among MOH and MOA employees in Tabora
 Variables:
 Smokers
 in MOH
 in MOA
 Definition:
 A smoker is someone who, at the time of the survey, smokes
cigarettes or any other tobacco products either daily or
occasionally.
 daily: smokes any tobacco product ≥ 1 / day
 occassionally: smokes, but not every day

Identify and define study variables 

Establish plan of analysis (POA)
 How will the key variable(s) be measured
to meet the objectives?
 Proportion, mean, ratio, ...
 Level of measurement influences
 questionnaire design
 type of statistical analysis
 What additional variables will need to be
measured?
 How will the results be used in the analysis
(dummy tables)?

Plan of analysis
To measure and compare the prevalence of smoking among MOH and MOA
staff in Tabora
Variables obtained from all study participants
Number (%) of persons smoking
- daily
- occasionally
Number (%) of persons smoking cigarettes, pipefuls of tobacco, or cigars /
cigarillos
Number (%) of nonsmokers who are
- neversmokers
- exsmokers (and time since quitting)
Variable obtained from smokers
Average number of cigarettes, pipefuls of tobacco, cigars / cigarillos smoked per
day or week

Baseline characteristic of study population in UNHCR and UN
--------------------------------------------------------------------------------------------------------
MOH MOA
N = X N = Y
--------------------------------------------------------------------------------------------------------
number (percent)
Age*
30 years or younger
31 - 45 years
46 years or older
Female
Work status
Staff member
Short Term Professional
Short Term Consultant
Others
Dummy table - describe sample

Prevalence of smoking among staff of UNHCR and UN
------------------------------------------------------------------------------------------------------
MOH MOA
N = X N = Y
------------------------------------------------------------------------------------------------------
number, percent
Smokers
smoking daily
smoking occasionally
total n = x n = y
Nonsmokers
neversmokers
exsmokers
quit ≤ 1 month ago
quit 2-6 months ago
Dummy table - describe variables

Establish plan of analysis 

Estimating sample size
• Indicate expected frequency of factor of interest
• Decide on desired precision of the estimate
• Decide on acceptable risk that estimate will fall
outside its real population value
• Adjust for estimated design effect
• Adjust for expected response rate
• (Adjust for population size, if sample > 10% of
target population)

Methods used in probability samples
 Simple random sampling
 Systematic sampling
 Stratified sampling
 Cluster sampling
 Multi-stage sampling

Determine sample size 
Choose sampling method 

Method of data collection
 Personal interview
 Telephone
 Self administered
 Questionnaire
 Mail
 Computer direct
 E-mail

What method to choose?
 Nature of questions
 What is the research question to be
addressed?
 Convenience
 Which method is most readily available?
 Cost
 How much money is available?

Personal interviews
 “face-to-face”
 Advantages
 population coverage
 can clarify issues, allows complex questions
 longer interviews tolerated (2-3 hours)
 usually high response rate (~90%)
 Disadvantages
 expensive
 slow
 intrusive
 requires interviewing skills
 interviewer bias

Self-administered questionnaires
(mail or hand distributed)
 Advantages
 cheap
 no interviewer bias
 long list of response categories possible
 sensitive questions
 Disadvantages
 slow
 does not allow clarification
 poor response rate (particularly in populations
of lower education and literacy level)
 population coverage?
 who completes questionnaire??

Telephone surveys
 “Averages” advantages and disadvantages of
personal and mail surveys
 Advantages
 relatively fast
 sampling through random digit dialling
 Disadvantages
 need telephone
 problem with answering machines, mobile
telephones
 no one at home during the day. Limits calling time
to a window of 6-9 p.m.
 interview time < 20 minutes

Determine sample size 
Choose sampling method 
Decide on mode of data collection 

Preparing for data analysis
 Remember “GIGO” principle
 “Garbage in - garbage out”
 The computer cannot make data better than
they are
 Coding
 Data entry
 Data cleaning and editing

Data analysis
 Type of data analysis is dependent on the
of outcome variable
 Continuous variables
 Categorical variables

Points about survey data analysis
 Target population in a survey often finite
 Most software packages assume:
 infinite population sizes
 sampling with replacement
 both conditions often not upheld in surveys
 Software specially developed for surveys
 SUDAAN (Survey data analysis)
 Stata
 Epi Info CSurvey

Ethical issues
 Informed consent
 purpose of study explained?
 consent given by individual
to be included?
 Confidentiality
 an individual should not be identifiable
 Service provision

Response rate
 Dependent on
 length and mode of interview
 educational level
 perceived threat of questions
 “warm” versus “cold” interviews
 warm = interviewee has prior knowledge about
survey
 Non-responders are almost always
different from responders
 To reduce non-response bias
 follow up non-responders at least once

Use of results
 Biggest fault in survey design: Results are
not disseminated or used.
 Why was the suryey carried out in the first
place?

Malimu intro to surveys

More Related Content

What's hot (15)

Viewers also liked (18)

Similar to Malimu intro to surveys (20)

More from Miharbi Ignasm (20)

Recently uploaded (20)

Malimu intro to surveys

Editor's Notes