Sept. 25, 2013
Please submit by Oct. 4th.
1. “OccupSat”
There are four types of occupations and workers from different occupations responded their satisfaction at their work. The final score came from responses of 18 questions and summed up. Higher numbers mean higher satisfaction.
1) Have a cross-tab for occupation and satisfaction
2) What are the means and standard deviations of the satisfaction for all workers surveyed? And, for each occupation?
3) What conclusions can you draw regarding the occupation satisfaction?
2. “Ready to Eat Cereals”
The raw data on ready-to-eat cereals collected by Roberts and Lattin (used as a sample in the chapter) are available in the file RTE_CEREAL. The file contains 27 variables defined as follows:
Column
Variable
Column
Variable
Column
Variable
1
Subject ID
10
Energy
19
Crisp
2
Cereal ID
11
Fun
20
Regular
3
Filling
12
Kids
21
Sugar
4
Natural
13
Soggy
22
Fruit
5
Fibre
14
Economical
23
Process
6
Sweet
15
Health
24
Quality
7
Easy
16
Family
25
Treat
8
Salt
17
Calories
26
Boring
9
Satisfying
18
Plain
27
Nutritious
Details for Cereal ID
Brand No.
Brand
Brand No.
Brand
Brand No.
Brand
1
All Bran
5
Komplete
9
Special K
2
Cerola Muesli
6
NutriGrain
10
Sustain
3
Just Right
7
Purina Muesli
11
Vitabrit
4
Kellogg’s Corn Flakes
8
Rice Bubbles
12
Weetbix
a) Do your own principal component analyses. See how many components can capture at least 60% of the original variance.
b) Conduct your own factor analysis. Try extracting and rotating four or five factors and see whether the results are different. [Note: The textbook provides results for results with four factors.]
3. Food research
A doctoral student in food research has a data set with 10 variables. She is concerned that 10 variables are too many for a subsequent analysis she needs to run. Ideally, she is hoping to somehow reduce her data set—perhaps by omitting