Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

Influence of positive and negative levels on the utilities and impacts of a attribute


I did a CBC with lighthouse Studio to explore relevant determinants for the employer choice.
There were for each attribute negative and positive levels (f.e. low salary and high salary)
I would like to find out, if negative levels have a bigger impact on the employer choice than positive levels.
As the utilities for one levels is the negative utilitiy of the other level for attributes with two levels, I am not sure if there is a way to find out if negative or the positive level has a bigger impact on the choice.

In this context, i would also love to find out if the importance for an attribute ist low, because the positive level has no big positive impact or because the negative level has a big negative impact.

Thank you and best regards,
asked May 22, 2019 by bea007 (255 points)

1 Answer

+1 vote
Positive and negative utilities result because we use effects coding for each attribute, which centers utilities at zero.  So you will ALWAYS get some positive and some negative utilities.  

The best way to measure impact is by running sensitivity analysis via simulations.  For example, for one of the concepts in your simulation hold everything constant in three simulation runs.  In one run let the salary be low, in one medium and in one high.  The simulation will show employer choice probability by salary level.  So if your shares for low/medium/high salary are 20%, 25% and 40%, you might conclude that high salary has more positive impact than low salary has negative impact.  Or if the salaries were instead 2%, 25% and 30%, you might conclude that low salaries have more negative impact than high salaries have low impact.  Showing the sensitivity analysis on a graph would make these relations even clearer.   

I should note that you don't have to run these simulations manually - the Choice Simulator available in our software gives you an easy way to run sensitivity analysis.
answered May 22, 2019 by Keith Chrzan Platinum Sawtooth Software, Inc. (103,225 points)
Thank you very much for your answer.
Unfortunately my attributes only have two levels, so this method does not work to identify if a positive  or negative level has more impact within an attribute.
Do you have another idea, how I can conclude if positive  or negative levels have more impact?

Thank you very much,
In your case positive and negative effects can only be two sides of the same coin.  Since there is no neutral baseline, the positive effect is the difference in share between the positive and negative endpoints.  And the negative effect is just the reverse of that.  An answer to your question would require that you had designed in a neutral baseline level to which to compare positive and negative outcomes.