Hello..
Example of my question is the following statement in a survey:
"I encourage my friends to buy this brand"
In the analysis stage, can I define this statement to represent (measure) more than one variable, for example Brand Loyalty, and Customer Satisfaction ??
Thank you
A question evaluates only one specific factor, so it is only one element of the mentioned construct, you have put 2 variables that would be evaluated with different tests, but you can make a study through factorial analysis if you want to integrate both factors through one instrument if you present a theoretical model. Although it is a controversial topic there are studies made by means of a single item that evaluates a construct, most factors have a main indicator, if you use the two golden indicators of your variables you can evaluate with 1 indicator per variable, but as you mention it it is a questionable topic methodologically, but it has several studies made.
Brand Loyalty and Customer Satisfaction are well known constructs, so I suggest you to not re-invent the wheel, but to use items (questions) from validated scales.
Having the same question in 2 different scales that measure 2 different constructs in a model is hard to imagine as valid.
Thank you Kenan Mahmutovic and Cristian Ramos-Vera ...
The problem I am facing is the length of questionnaire and repeatability of look-alike questions from respondents point of view.
For example and in relation to your suggestion Kenan to no reinvent the wheel, I will have to ask ... I encourage my friends to buy this brand and put another similar question I suggest this brand to my friends in order to measure the two different variables, i.e. Brand Loyalty, and Customer Satisfaction.
From pilot testing the survey I got several remarks about the similarity and repeatability of questions that look alike participants, who are@ unlikely to understand the rationale for the repetition of questions.
I agree with the two responses. It is always striking a balance between having the minimum number of questions and getting relevant data. For well-established constructs as brand loyalty and customer satisfaction, I am sure you will have a hard time convincing your reviewers why you need to come up with new measurement instruments. As you pointed out in your example, it is worth asking yourself how you are going to explain whether the behaviour is a result of satisfaction or loyalty?
Totally agree that you need to keep the questionnaire as short and concise as you can. (And on re-reading my response not entirely sure I am being helpful in regard to length.) However, it would seem that you have not analyzed / broken down the constructs as thoroughly as you could. i.e. different variables make up the constructs.
Early in the questionnaire you should test awareness of brands … maybe it is a simple factor that people are not aware of the product/brand.
You may also be interested in the pricing differentials. i.e. if brand X is the premium brand you could look at how much cheaper the closest competitor has to be to switch purchase. Maybe brand image is not as important as price!
Then presuming you have established the buying factors either by groups or a literature review, you would position the different brands on the scales that describe them. (e.g. for a food product the two ends of a semantic differential could be extremely pleasant taste / not at all pleasant tasting)
If I was measuring brand loyalty, I would consider several indicators and where applicable have a probe: Why do you say that?
(In the following I am trying to think of the type of questions that would be useful. Assume brand X is the focus…)
1) Purchase profile:
Which brands of product are you aware of? (spontaneous)
Show list Which of these brands of product have you ever used?
If uses target brand: How many out of the last 10 times that you bought this product did you buy brand X ?... or could rather ask this Considering this list of Brands – of your last 10 purchases of this product how many were Brand X, Brand Y, Brand Z and so on.)
2) Attitude to brands: For the top three brands: What words would you use to describe brand X? Brand Y? Brand Z?
Is Brand X better/the same/ not as good as other brands?
3) Usage time: For approximately how long have you used Brand X? OR When did you start using brand X?
4) Substitution: If Brand X was not available would you
a) Not buy the product at all
b) Go to an outlet where you could buy brand X
c) Buy another brand
d) Buy another product type
e) Other (state)
5) Recommendation: How likely are you to recommend brand X to your friends? (very likely / likely / neither likely, nor unlikely / unlikely / very unlikely)?
6) Unique benefits: What benefits does brand X have that the other brands do not have?
7) Comparison of brand features --- Use Likert scales to measure the importance of product features. Then use semantic differentials to get the respondents to rate the different brands on the product features.
You can see that in some of the above one question can provide insight into different aspects of the various brands relative positions….
- Badges
- Science topic
- Similar topics
- Statistics
- Survey