Exercises for Chapter 1

2.1. In Example 1 you met a summated scale intended to measure peoples' 'general experience/competence regarding preparation of fish'. In planning the same project, the following items were suggested for measuring 'experience/competence regarding cooking of fish dishes':

 

43. When I prepare fish dishes, I prefer to use a cookery book.
44. When I prepare fish dishes, I often think I fail.
45. I like to prepare complicated fish dishes (NB: the notion of "complicated" to be interpreted by the respondent him/herself).
46. When I prepare fish dishes, I prefer dishes that are easily and quickly cooked.
47. Preparing fish dishes gives me the possibility of using my abilities as a cook.
48. I have many different dishes to choose from when serving fish.
49. When I am to prepare fish dishes, I stick to dishes that I know well so that I am sure to succeed.
50. I think I know too few fish dishes to alternate between.

Do an item analysis (reliability analysis) on this scale.

Click here for the data set

3.1. Perform a principal component analysis and an exploratory factor analysis on the data used in Exercise 2.1.

5.1. The structural model shown below (excl. error terms) has been proposed as basis for a mapping of the factors affecting young peoples' loyalty towards their bank. The model is inspired by 'The European Customer Satisfaction Index'. Estimate their model using summated scales.

The latent variables are measured as follows:
Name of bank: Item 1
Quality: Items 2-7
Trust: Items 8-14
Image: Items 15-24
Satisfaction: Items 25-30

Click here for the data set.

This exercise is based on a student project from my first year graduate class in research methodology at Aarhus School of Business, autumn 2006. I would like to thank the four business students Johnny Heidemann Jensen, Mads Ronøe Hansen, Ole Primholdt Christensen and Mikael Martinussen for allowing me to use their data.

5.2. The (structural) model shown below (excl. error terms) has been proposed as a basis for mapping the various factors affecting students' intentions of buying food via the Internet. The model is inspired by 'The Theory of Planned Behavior' (Ajzen 1985; Ajzen 1991) as modified by Ramus and Grunert (2004). Estimate their model using summated scales.

The latent variables are measured as follows:
Demand for service: Q17 –Q25
Willingness to spend money on food: Q1 – Q8
Willingness to run a risk: Q9 – Q16
Busy life style: Q34 – Q41
Demand for convenience: Q26 –33
Attitude: Q49 –Q57
Needs: Q42 –Q48
Intention to buy: Q58 –Q60

Click here for the data set

This exercise is based on a student project from my first year graduate class in research methodology at Aarhus School of Business, autumn 2006. I would like to thank the four business students Louise Tüxen, Rune Klemmensen, Martin Bjerre Hansen and Anne Høj Pedersen for allowing me to use their data.

Ajzen, I. (1985), "From intentions to actions: A theory of planned behavior," in Action-control: From cognition to behavior, J. Kuhl and J. Beckman, Ed. Heidelberg: Springer.

---- (1991), "The theory of planned behavior: ," Organizational Behavior and Human Decision Processes, 50, 171-211.

Ramus, K. and K. G. Grunert (2004), Consumers' willingness to buy food via the Internet A review of the literature and a model for future research. Aarhus: Aarhus School of Business.

6.1. Judge the qualities of the scale from Example 2.1 by using confirmatory factor analysis, and compare the results of your analysis with the results from traditional reliability analysis (Example 2.1) and exploratory factor analysis (Example 3.3).

6.2. Do a confirmative factor analysis using the data from Exercise 2.1 and compare the results.

6.3. Do a confirmative factor analysis on the data used in Example 3.4.

7.1. Re-analyze the model from Exercise 5.1, using the techniques from this chapter.

7.2. Re-analyze the model from Exercise 5.2, using the techniques from this chapter.

8.1. As mentioned in Example 7.2, Kirkegaard et al. also included a sample of young people in their study. Examine the equivalence of their measuring instruments across the two samples (young and elder people).

Click here for the data set

8.2. To what extent is does the causal model from example 7.2 fit the young sample?

9.1. At the end of Example 7.2 I pointed out that although indirect and total effects could be estimated by 'ordinary' ML (or any of the other estimation methods available), these methods would not report standard errors and P-values for such effects. We have to fall back on bootstrapping for testing.

Now that you have had an opportunity to look at the raw data, you will have observed that most manifest variables are severely skewed (luckily they are all skewed to the same side, which makes the problem a bit less serious) and that bootstrapping perhaps would solve that problem too. So, estimate the model from Example 7.2 by bootstrapping and test for indirect and total effects.

Click here for the data set

9.2. Experiment with the data (click here for the data set). Try various ways of coping with missing values and with the non-normality of the data – e.g. compare various transformations and estimation methods. You are also welcome to check for outliers and other data problems.

Exercises for Chapter 10