PIR1600 WEEK 10: SURVEYS IN SOCIAL SCIENCES

1
PIR1600WEEK 10 SURVEYS IN SOCIAL SCIENCES

• SECKIN BARIS GULMEZ
• s.b.gulmez_at_rhul.ac.uk
• www.pir1600.wordpress.com

2
SURVEYS

• COMPULSORY READINGS
• Peter Burnham, Karin Gilland Lutz, Wyn Grant and
  Zig Layton-Henry (2008). Research Methods in
  Politics (Palgrave Macmillan). Chapter 4
• Oliver Heath, Research Methods in IR, chapter 8,
  2011.

3
QUESTIONS

• What is the aim of surveys? (Burnham et al, 2008)
• Obtain accurate information about the target
  group/population (Burnham et al, 2008 97)
• How to do it?
• Obtain representative sample of the group
• Use the sample for generalizations about the
  group
• (Burnham et al, 2008 97)

4
QUESTIONS

• What types of information does a good survey aim
  to gather? (Heath, 2011)
• 1- Descriptive information
• (e.g. the level of support for democracy)
• 2- Analytical information (Hypothesis or theory
  testing)
• (e.g. the reasons behind support for democracy)
• (Heath, 2011 2)

5
DESCRIPTIVE INFORMATION
6
ANALYTICAL INFORMATION
7
SAMPLING

• What are the types of taking the sample of
  population/target group? (Burnham et al, 2008)
• Random Sampling (Lottery?)
• Non-Random Sampling

8
Random Sampling

• What are the types of random sampling?
• 1- Systematic RS
• choosing the 5th, 10th, 15th , 20th of the group
  etc .
• Problems?
• Problems of representation all male or female,
  all under-educated, under-aged etc
• 2-Stratified RS
• Divide the sample into homogeneous groups/strata
• It is a better way for increasing the
  representativeness of the sample.

9
Random Sampling

• 3- Multi-stage Sampling
• Used for large national surveys.
• Divide country into constituencies and then take
  the sampling of each constituency.
• -Difference from stratified sampling?
• Larger target group
• Convenience not homogeneity

10
Random Sampling

• 4-Cluster Sampling
• Similar to Multi-stage sampling
• Difference?
• After dividing the sample into constituencies,
  the researcher chooses a village/group from each
  constituency and interview them.
• Problems with this sampling?
• Issue of representativeness

11
NON-Random Sampling

• 1-Quota Sampling
• Choose the sample of the population according to
  certain quotas age, sex, profession in order to
  proportionately represent the whole target group.
• X amount of man, x amount of women, y amount of
  working class women, y amount of housewives etc
• 2- Snowball Sampling
• Reaching more and more respondents via the
  suggestions of other respondents.

12
DESIGNING THE QUESTIONNAIRE

• What are the important issues the surveyor needs
  to be aware of?
• Clarity of the questions
• Questions prone to be misunderstood by respondent
• Leading questions
• Logical sequence of the question (general to
  specific)
• Length of the questions
• Language

13
EXAMPLES TO DISCUSS

• How often do you read?
• What is your income?
• How often do you book your holidays on the
  internet?
• Do you believe Iran has nuclear weapons and what
  should the US do about that?

14
EXAMPLES TO DISCUSS

• Dont you think W. Bush was more of a warmonger
  than a President?
• Do you think that Turkey should abrogate the
  Article 301 of the Turkish Penal Code?
• (which stipulates that A person who publicly
  denigrates Turkish nation, the Republic or the
  Grand National Assembly of Turkey, shall be
  punishable by imprisonment)

15
QUESTION TYPES

• Open ended question?? Problems??
• Closed questions?? Problems??
• Semi-structured questions??

16
EXAMPLES

• Do you think there is an opposition in your
  country against Turkeys EU membership?
• 
  .
• Do you think there is an opposition in your
  country against Turkeys EU membership?
• Yes
• No

17
EXAMPLES

• Do you think there is an opposition in your
  country against Turkeys EU membership?
• There is a strong opposition
• There is an opposition but not strong
• There is no such an opposition
• Dont know
• Do you think there is an opposition in your
  country against Turkeys EU membership?
• There is a strong opposition
• There is an opposition but not strong
• There is no such an opposition
• Dont know
• Other (Please specify)

18
MEASUREMENT ERRORS

• Comprehension problems
• Recall problems
• Deliberate misreporting
• Question order effects
• Question wording effects
• Leading questions
• Response format

19
FINAL QUESTIONS

• What is self-administered questionnaire??
• No interview
• Questions sent via post, email or through a
  webpage
• Strengths and Weaknesses??

20
TERM ENDS..
MERRY CHRISTMAS AND/OR HAPPY NEW YEAR