29th October 2018- Research Methods
Research Methods- Surveys:
What is a Survey?
- Warwick and Lininger (1975): ‘A survey is a method of collecting information about a human
population in which direct contact is made with the units of study (individuals, organizations,
communities, etc) through such systematic means as questionnaires and interview
schedules'
- Fink and Kosecoff (1985): ‘Survey research employs questionnaires and interviews to ask
people to provide information about themselves – their attitudes and beliefs, demographics
(age, gender, income, marital status, and so on) and other facts, and past or intended future
behaviours’
- In all surveys, the type of data obtained from participants = self report.
Different Types of Surveys:
- Descriptive Surveys= main aim is to ‘count’ some variable(s) to indicate the proportion of a
population that have a certain opinion, characteristic, etc.
- Analytic Surveys= main aim is to explore associations between particular variables, i.e.
correlations and regression.
Descriptive Surveys:
- Study almost any behaviour, e.g. consumer behaviour, attitudes, psychological effects of
many types of experiences, etc.
- Surveys are often used to study very ‘personal’ issues, e.g. human sexual activity,
psychopathologies.
- Obtain data where experiments or direct observations would be tricky.
- Allow researchers to compare ‘groups’.
- E.g. Hochard et al. (in prep.) surveyed 783 individuals recovering from Eating Disorders
about their engagement in self harm:
To test hypothesis that depending on stage of recovery, self-harm rates would differ.
Found 91.8% of sample overall reported having previously self-harmed and 65.8%
made a suicide attempt.
Why Perform a Descriptive Survey?
3 reasons to perform a descriptive survey (Fink and Kosecoff, 1985):
1) A policy needs to be set or a programme planned- information e.g. SU survey students to
know what type of beer they want.
2) There is a need to evaluate the effectiveness of programmes to change peoples knowledge,
attitudes, health, welfare, etc. e.g. prior to, and then following, an AIDS awareness course, a
survey is carried out to assess participants’ willingness to use a condom.
3) To help answer a research question. E.g. survey to assess the level of bullying among high
school pupils.
Analytic Surveys:
- Data from analytic survey provides insights into cause and effect relationships, but always be
cautious about accepting conclusion about cause and effect relationships based on data
from (cross-sectional) surveys.
- E.g. my survey revealed a significant positive correlation between amount of alcohol
consumed and level of violence. But can we conclude that alcohol CAUSES violence? No.
Other Surveys are Concerned with Prediction:
- e.g. opinion polls to forecast election results.
1
, 29th October 2018- Research Methods
- but opinion surveys can go ‘spectacularly’ wrong, e.g. Brexit, Trump election. Because:
The sample might not be representative of the population.
Respondents might change their minds
Respondents might deliberately give the wrong information.
Methods of Data Collection in Surveys:
- Multiple options are open to researchers.
- The top 5 most common methods are:
1) Face-to-face interviews
2) Telephone interviews
3) Questionnaires
4) Postal questionnaires
5) Internet
Face-To-Face Interviews:
Advantages:
- People are more likely to agree to participate
- People are more likely to agree to longer interviews
- Interviewer can ‘probe’ questions to follow up things of interest that emerge in the
interview (e.g. can you explain that further? What do you mean when you say…?)
Disadvantages:
- Tend to be costly
- Data collection period is likely to be relatively long- time consuming
- Some issues may be too personal/embarrassing for face-to-face discussion
Telephone Interviews:
Advantages:
- Cheaper than face-to-face interviews
- Random-digit dialling enables random sampling (or does it?)
- Higher response rate than postal surveys
Disadvantages:
- Sampling limitations – those without telephones excluded
- Cannot use visual aids to generate responses
- Response rate lower than face-to-face interviews
Questionnaires:
Advantages:
- Respondents don’t have to ‘share’ answer with interviewer – may be more honest
- Can ask questions with long/complex response categories
- Can use visual aids to generate responses
Disadvantages:
- Open questions tend not to be answered in depth
- Questions cannot be explained to respondents
- Respondents need good reading and writing skills
Postal Questionnaires:
Advantages:
- Can be used to reach people not available for telephone of face-to-face surveys
- Respondents have more time to answer
- Less expensive than face-to-face interviews
Disadvantages:
2
Research Methods- Surveys:
What is a Survey?
- Warwick and Lininger (1975): ‘A survey is a method of collecting information about a human
population in which direct contact is made with the units of study (individuals, organizations,
communities, etc) through such systematic means as questionnaires and interview
schedules'
- Fink and Kosecoff (1985): ‘Survey research employs questionnaires and interviews to ask
people to provide information about themselves – their attitudes and beliefs, demographics
(age, gender, income, marital status, and so on) and other facts, and past or intended future
behaviours’
- In all surveys, the type of data obtained from participants = self report.
Different Types of Surveys:
- Descriptive Surveys= main aim is to ‘count’ some variable(s) to indicate the proportion of a
population that have a certain opinion, characteristic, etc.
- Analytic Surveys= main aim is to explore associations between particular variables, i.e.
correlations and regression.
Descriptive Surveys:
- Study almost any behaviour, e.g. consumer behaviour, attitudes, psychological effects of
many types of experiences, etc.
- Surveys are often used to study very ‘personal’ issues, e.g. human sexual activity,
psychopathologies.
- Obtain data where experiments or direct observations would be tricky.
- Allow researchers to compare ‘groups’.
- E.g. Hochard et al. (in prep.) surveyed 783 individuals recovering from Eating Disorders
about their engagement in self harm:
To test hypothesis that depending on stage of recovery, self-harm rates would differ.
Found 91.8% of sample overall reported having previously self-harmed and 65.8%
made a suicide attempt.
Why Perform a Descriptive Survey?
3 reasons to perform a descriptive survey (Fink and Kosecoff, 1985):
1) A policy needs to be set or a programme planned- information e.g. SU survey students to
know what type of beer they want.
2) There is a need to evaluate the effectiveness of programmes to change peoples knowledge,
attitudes, health, welfare, etc. e.g. prior to, and then following, an AIDS awareness course, a
survey is carried out to assess participants’ willingness to use a condom.
3) To help answer a research question. E.g. survey to assess the level of bullying among high
school pupils.
Analytic Surveys:
- Data from analytic survey provides insights into cause and effect relationships, but always be
cautious about accepting conclusion about cause and effect relationships based on data
from (cross-sectional) surveys.
- E.g. my survey revealed a significant positive correlation between amount of alcohol
consumed and level of violence. But can we conclude that alcohol CAUSES violence? No.
Other Surveys are Concerned with Prediction:
- e.g. opinion polls to forecast election results.
1
, 29th October 2018- Research Methods
- but opinion surveys can go ‘spectacularly’ wrong, e.g. Brexit, Trump election. Because:
The sample might not be representative of the population.
Respondents might change their minds
Respondents might deliberately give the wrong information.
Methods of Data Collection in Surveys:
- Multiple options are open to researchers.
- The top 5 most common methods are:
1) Face-to-face interviews
2) Telephone interviews
3) Questionnaires
4) Postal questionnaires
5) Internet
Face-To-Face Interviews:
Advantages:
- People are more likely to agree to participate
- People are more likely to agree to longer interviews
- Interviewer can ‘probe’ questions to follow up things of interest that emerge in the
interview (e.g. can you explain that further? What do you mean when you say…?)
Disadvantages:
- Tend to be costly
- Data collection period is likely to be relatively long- time consuming
- Some issues may be too personal/embarrassing for face-to-face discussion
Telephone Interviews:
Advantages:
- Cheaper than face-to-face interviews
- Random-digit dialling enables random sampling (or does it?)
- Higher response rate than postal surveys
Disadvantages:
- Sampling limitations – those without telephones excluded
- Cannot use visual aids to generate responses
- Response rate lower than face-to-face interviews
Questionnaires:
Advantages:
- Respondents don’t have to ‘share’ answer with interviewer – may be more honest
- Can ask questions with long/complex response categories
- Can use visual aids to generate responses
Disadvantages:
- Open questions tend not to be answered in depth
- Questions cannot be explained to respondents
- Respondents need good reading and writing skills
Postal Questionnaires:
Advantages:
- Can be used to reach people not available for telephone of face-to-face surveys
- Respondents have more time to answer
- Less expensive than face-to-face interviews
Disadvantages:
2
