1. Web-based online surveys are good because ______.
- you can use your list of friends and contacts to obtain good quality data
- they can be customised for each respondent
- you can control the quality of answers from respondents
- replies are more truthful
2. Which of the following is true about benefits of alternative survey strategies?
- A large sample is always better than a small sample.
- The internet means that interviewing face to face is no longer necessary.
- Building a good relationship with respondents will increase their willingness to give high quality answers.
- Interviews are too expensive.
3. In order to increase response rates in surveys, which of the following is true?
- People will only reply if they think they will benefit personally.
- Treating study participants with respect is important.
- It is necessary to offer something in order to get people to reply.
- Students can never expect to get a good response rate from their surveys.
4. Which of the following is true about observational data?
- How people behave gives important insights into how they think and feel.
- It is easy to capture every aspect of behaviour using observational methods.
- Recording behaviour means that it is not necessary to think in advance about your research hypotheses.
- Recording people without them being aware gives the best data.
5. Which is incorrect about secondary databases?
- It is not necessary to check the quality of data from secondary databases.
- Archives are valuable sources of data which could not be obtained in any other way.
- Secondary data can be used for purposes which are different from the reason why they were collected in the first place.
- It is important to justify how study variables are derived from secondary data sources.
6. In designing structured questions for a survey, which of the following is true?
- Simple questions are generally the best.
- Putting questions in a random order will encourage respondents to think hard about each question.
- It should always be easy for respondents to know what is your preferred answer to a survey question.
- Questions from published scales should always be used with exactly the same wording as in the original article.
7. Gender is measured on an ordered category scale.
- both true and false
8. Which of the following is true of continuous measurement scales?
- Likert scales can never be treated as truly continuous.
- Continuous scales make no sense unless they have a true zero point.
- It is appropriate to calculate ratios for variables measured on an interval scale.
- Ordered category scales can be treated as approximations of continuous scales.
9. In defining a measurement model, which of the following is NOT true?
- A measurement model specifies the relationship between observed variables and a hypothesised latent construct.
- More than one measured variable is usually needed in order to capture the richness of constructs that researchers are interested in.
- The relationship between an observed variable and a latent construct is shown by a correlation coefficient.
- Observed variables which measure the same latent variable will usually be correlated.
10. In fitting measurement models, which of the following is true?
- Factor analysis models specify the relationship between observed and latent variables.
- It is better not to have a specific hypothesised structure in mind when fitting a measurement model.
- Output from CFA always tells the researcher the names of latent variables.
- Observed variables are assumed to be causes of latent variables.
11. A written transcript of a conversation or an interview contains all the relevant information.