L11 - Conjoint measurement analysis Flashcards

1
Q

Goal of the conjoint measurement analysis

A

Identify and measure the attributes and attribute levels that people care most about ( that have the strongest impact on preferences)
–> decompositional method

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is the principle of the conjoint analysis?

A

Create options presenting combinations of different attribute levels and measure the preferences for different options; then infer the influence of the different attributes and attribute levels.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Example conjoint analysis

A

Estimating the influence of individual product features on preferences.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What are the steps in the conjoint measurement analysis?

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

3 key assumptions of conjoint analysis

A
  • preference is based on a combination of all attributes (they are CONsidered JOINTly)
  • attributes can compensate each other
  • the attributes are independent
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Is relevance important when selecting the attributes?

A

Yes. The attribute must be relevant for people’s preferences (-> identification of attributes using focus groups)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Is control important when selecting the attributes?

A

Only choose attributes and levels that can be modified and realized during product design.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Is independence important when selecting attributes?

A

Yes the attributes should be independent.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Is compensatory important when selecting the attribute?

A

Yes. The attributes can compensate each other (e.g. a reduction of the calorie content can be compensated by an improvement in taste)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Is an exclusion criteria allowed when selecting attributes?

A

No. The attribute levels should not be exclusion criteria (“knock-off criteria”).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Do the number of attributes and attribute levels have to be limited?

A

Yes.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

How to calculate the number of profiles?

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is a full factorial design?

A

When you list all the profiles (combinations of attributes and attribute levels). This can result in too many options for the customer.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

How can you reduce the number of combinations of the profiles?

A

Fractional design with latin square. -> You can maximize information without considering every option.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What is the criteria you have to fulfill so that you can apply the fractional design?

A

Only if no interaction between the attributes is to be expected.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is the latin square restricted to?

A

Situations with 3 attributes and 3 levels.

17
Q

What is the principle of the latin square?

A

Each attribute level is combined once with each other attribute levels of the other attributes.

18
Q

What is another fractional design method?

A

Orthogonal design?

19
Q

When do you use the orthogonal design?

A

When you have more than 3 attributes and levels.

20
Q

What is the assumption of the orthogonal design?

A

That no interactions between attributes are to be expected.

21
Q

What is the goal of the orthogonal design?

A

That the attributes are uncorrelated. It is based on a simulation methods.

22
Q

Ordinal methods for preference assessment

A
23
Q

Metric methods for preference assessment

A
24
Q

What is the ranking method?

A

Please rank the 9 options according to how attractive you find them.

25
Q

What is pair comparison?

A

Please indicate which option you find more attractive

26
Q

What is the rating scale?

A

Please indicate how attractive you find option A on a scale from 1 to 10

27
Q

What is the dollar metric?

A

How much more would you pay for option A relative to option B
or
How much would you pay for option A

28
Q

Constant sum method

A

Distribute 100 points across the options such that a more strongly preferred option receives proportionally more points.

29
Q

What is the coding matrix about?

A

The presence of absence of attribute levels

30
Q

What is the part-worth matrix?

A

How important attribute levels are for a person.

31
Q

How can you use the results of a conjoint analysis?

A
  • It can have implications for existing products
  • simulation of preferences for novel products
32
Q

Predicting utility of other profiles

A

4.75-1(-1)+1.7-1-.70+-21-.51-.25-1

33
Q

What do we estimate with the logit model in the choice-based conjont analysis?

A

probability that a particular option is chosen.