using weights
Great question! If it isn't implemented in this release that would be a wonderful feature for the next.
But I'm not sure about boostrapping with a variable weight. The variable weight, for example, might be to make the data more representative (ie., your data is skewed sucht that there are too many female respondents). Those weights may not be relevant in the boostrapped samples.
But I'm not sure about boostrapping with a variable weight. The variable weight, for example, might be to make the data more representative (ie., your data is skewed sucht that there are too many female respondents). Those weights may not be relevant in the boostrapped samples.
John J. Sailors, PhD
Associate Professor of Marketing
The University of St. Thomas
Opus College of Business
Minneapolis, MN
Associate Professor of Marketing
The University of St. Thomas
Opus College of Business
Minneapolis, MN
Let's imagine that we have data from owners of products in a particular category, Brand A, B, and C. In our data, the distribution of ownership is as follows:
Brand A 15%
Brand B 50%
Brand C 35%
However, the brands have the following market shares:
Brand A 25%
Brand B 35%
Brand C 40%
Especially in applied research it is common to apply weights to the observations so that the weighted shares in the data set equal the actual market shares.
That's what I meant, at least, can't speak for anyone else.
Brand A 15%
Brand B 50%
Brand C 35%
However, the brands have the following market shares:
Brand A 25%
Brand B 35%
Brand C 40%
Especially in applied research it is common to apply weights to the observations so that the weighted shares in the data set equal the actual market shares.
That's what I meant, at least, can't speak for anyone else.
John J. Sailors, PhD
Associate Professor of Marketing
The University of St. Thomas
Opus College of Business
Minneapolis, MN
Associate Professor of Marketing
The University of St. Thomas
Opus College of Business
Minneapolis, MN
- swende
- SmartPLS Developer
- Posts: 111
- Joined: Mon Sep 26, 2005 10:34 am
- Real name and title: Dipl. Wi-Inf. Sven Wende
- Location: Hamburg (Germany)
Thanks for the example. I think, now I got the message, too!
This is nothing, we have implemented right now, but it might be interesting for the next release.
Although I understand the problem, on second thoughts I´m not sure about how to realize this in the software.
Let me grab and simplify you example.
If I understand you right, your dataset looks like this:
Brand A Customer 1 2 3
Brand A Customer 3 4 3
Brand A Customer 2 2 2
Brand B Customer 3 5 6
So, the distribution of ownership is
Brand A 75%
Brand B 25%
How would you modify/weight your dataset, if the market shares were:
Brand A 50%
Brand B 50%
?
Hopefully I do not disgrace myself by asking this question, but I am not a practitioner in this topic! ;)
This is nothing, we have implemented right now, but it might be interesting for the next release.
Although I understand the problem, on second thoughts I´m not sure about how to realize this in the software.
Let me grab and simplify you example.
If I understand you right, your dataset looks like this:
Brand A Customer 1 2 3
Brand A Customer 3 4 3
Brand A Customer 2 2 2
Brand B Customer 3 5 6
So, the distribution of ownership is
Brand A 75%
Brand B 25%
How would you modify/weight your dataset, if the market shares were:
Brand A 50%
Brand B 50%
?
Hopefully I do not disgrace myself by asking this question, but I am not a practitioner in this topic! ;)
Sven Wende, CEO SmartPLS GmbH
In your example, where the distributions in the dataset are Brand A 75% and Brand B 25% but the actual market shares are 50% and 50% you would weight the observations as follows:
Each Brand A respondent would have a weight of .667 applied to it (so now each Brand A respondent is counted as .667 of a respondent) and each Brand B respondent would have a weight of 2 (so each Brand B respondent gets counted twice).
This will give you the 50% distribution across both brands that reflects the marketplace.
This is what weighting in both SPSS and SAS accomplishes. As I said, most often useful in applied research, I think.
Each Brand A respondent would have a weight of .667 applied to it (so now each Brand A respondent is counted as .667 of a respondent) and each Brand B respondent would have a weight of 2 (so each Brand B respondent gets counted twice).
This will give you the 50% distribution across both brands that reflects the marketplace.
This is what weighting in both SPSS and SAS accomplishes. As I said, most often useful in applied research, I think.
John J. Sailors, PhD
Associate Professor of Marketing
The University of St. Thomas
Opus College of Business
Minneapolis, MN
Associate Professor of Marketing
The University of St. Thomas
Opus College of Business
Minneapolis, MN