General population (methods)

toolkit material

In the general population and in young people, surveys are generally conducted using representative samples of these populations. The first table presents different types of sample construction.

A discussion of the advantages and disavantages of different modes of administration of the questionnaires in general population and youth surveys can be found in the second table.

Table 1. Selected features of available methods for collecting behavioural data among the general population and youth: Sample construction  

Design  Advantages   Disadvantages   Indications
Probability sampling Random sampling(1)
  • Provides population estimates
  • Allows statistical inference

Sampling frame maybe incomplete, out-of-date, etc

  • Preferred method
  • Common method today
  Random location
  • Provides population estimates
  • Allows statistical inference
   
  Random digit dialing (phone)
  • Provides population estimates
  • Allows statistical inference
  • Coverage limited to phone-owning population
  • Mobile phones make everything more complex (often no national list of owners)

Common method today

Representative sampling    Quota sampling 

Provides population pseudo-estimates through survey balancing, weighting, etc.

  • Does not allow statistical inference Survey balancing, weighting, etc. may hide selection biases
Common method today
Purposive or convenience sampling Convenience sample in venues, etc. May be only way to reach target population
  • Does not provide population estimates
  • Does not allow statistical inference
When no other solution is available (e.g. drop-out youth)

References (a)  Groves, Robert M., F.J. Fowler, Mick P. Couper, James M. Lepkowski, Eleanor Singer, and R. Tourangeau. 2004. Survey Methodology. New York: Wiley. Notes (1)  May be any type of simple or complex random sampling (multi-stage, etc.), including schools and/or classes as clusters for youth

Table 2. Selected features of available methods for collecting behavioural data among the general population and youth: Mode of administration

 

 

 

Mode Advantages Disadvantages Comments
Telephone interview  
  • Inexpensive
  • Allows for complex question flow 
  • Provides valid data
  • Increasing reluctance to answer phone interviews
  • Mobile phones make everything more complex
  • May need contact letter, etc. 
Most common method today
Face-to-face personal interview
  • Allows for complex question flow
  • Provides valid data
  • Most expensive
  • Access may be difficult
 
Paper questionnaire mailed by post
  • Inexpensive
  • Does not allow for complex question flow 
Most difficult to achieve good participation rate  
Emailed questionnaire 
  • Inexpensive
  • Does not allow for complex question flow
Difficult to achieve good participation rate   
Internet-based questionnaire
  • Inexpensive
  • Allows for complex question flow
Difficult to achieve good participation rate  
Paper questionnaire in situ (e.g. school, workplace)
  • Inexpensive
  • Does not allow for complex question flow

 

  Youth: most common method today
CASI, CAPI (1)  Allows for complex question flow    
Mixed modes 
  • Flexibility
  • May offer the best mix of features
   

 

 

Notes (1) CASI: Computer-assisted self interview. CAPI: Computer-assisted personal interview

Behavioural surveillance can be conducted in different methodological contexts. These contexts are discussed in the table below with their advantages and diadvantages. References of articles presenting examples of « good practice » are given.

Table. 3 Selected features of available methods for collecting data monitoring the general population and youth: Context of data collection process

 

 

Context Advantages Disadvantages Comments
In-depth Sex Survey (a)
  • Provides needed in-depth information on sexuality, etc.
  • Provides indicator data 
  • Expensive and cumbersome
  • Slow to implement
  • Not adequate for frequent monitoring 
Appropriate in early stages of the monitoring process
Specific KABP survey (b)  Provides indicator data     Appropriate for frequent monitoring
Transportable module in e.g. General Health Survey (c) (or Health barometers)
  • Provides indicator data
  • Context is appropriate
  • Inexpensive
  • Sustainable 
 
  • Appropriate for frequent monitoring
  • Probably the best context regarding sustainability, acceptability, breath of information
Transportable module in omnibus survey (d) 
  • Provides indicator data
  • Inexpensive
  • Sustainable
  • Flexible 
Context may threaten validity  Appropriate for frequent monitoring
Youth: module in general school survey (e.g. HBSC) (e) 
  • May provide indicator data
  • Context is adequate
  • Sustainable 
  • May not provide all needed indicator data
  • Limits age range accessed
 
Mixed contexts 

Flexible  

 

   

Examples (a)  A.Johnson, C.Mercer, B.Erens, A.Copas, S.McManus, K.Wellings, K.Fenton, C.Korovessis, W.Macdowall, K.Nanchahal. Sexual behaviour in Britain: partnerships, practices, and HIV risk behaviours. The Lancet, 2001 Dec 1;358(9296):1835-42. (b) Grémy I, Beltzer N. HIV risk and condom use in the adult heterosexual population in France between 1992 and 2001: return to the starting point? AIDS. 2004 Mar 26;18(5):805-9. (c) Robertson BJ. Sexual behaviour and risk of exposure to HIV among 18-25-year-olds in Scotland: assessing change 1988-1993. AIDS 1995;9(3):285-92 (d) Lader D. Contraception and sexual health 2006/07: a report on research using the National Statistics Omnibus Survey produced on behalf of the Information Centre for Health and Social Care. Newport: Office for National Statistics; 2007. (e) see www.hbsc.org.