Its difficult but you can give a try with "Nvivo" software

If you want to get there, as we say in Ireland, you shouldn't start from here.

If you want quantitative data, then the best way to get it is through quantitative methods. Laying the groundwork for your study is vital. Asking the right questions, and ensuring that the answers are distinct and represent all the potential responses requires pre-testing, often quite a lot of it.

Trying to generate quantitative data from interview data is really problematic. Take a simple case where you want to see how many people mentioned stress as a factor in the interview. The problem is that the people who don't mention stress may still think it's a factor, but the interview never got around to the point where they would have mentioned it. So the fact that a person doesn't bring something up doesn't mean it's not there, just that the interview didn't elicit it.

NVivo is for qualitative research, and if you have interview data, then qualitative methods are the way to go. You cannot map numbers onto interview data, but you can certainly extract useful insights.

You can prepare a checklist (like a newly started questionnaire based study) including some binary or categorical items. These items can be the key insights extracted form the interviews. Fill in your checklist based on the interview results.

In order to increase your cronbach's alpha, increase the number of questions (items) because unfortunately your variances are high (due to low sample size and low number of likert choices).

Then you can perform factor analysis and stepwise regression to extract the most important insights without nuisance.

Then perform structural equations modeling (SEM). You can define latent variables etc......

You can use SPSS, but you need Stata for SEM (or Amos instead).

I am not clear about whether the experts are the source of your data or whether their ratings are being used as the method for converting the data. Can you give us more information about what kind of data you are collecting and why you want to quantize it?

Hello everybody,

Thank you very much for your answers!

I want to interview experts to generate data (and gain an insight into which ESG factors are the most important). Based on that I would like to generate a weighting of the individual ESG factors (not just a ranking).

One classic way to get this kind of weighting is through a Delphi process. Often, the first round of a Delphi use qualitative data collection to help define the relevant factors, while subsequent rounds provide quantitative ratings on those factors.

I have attached an article that gives some examples of this method.

Viktoria Neudecker – the way to model this may be judgement analysis. In this method, you present experts with vignettes which are constructed so as to vary the parameters you are interested in. You then model their judgements are a function of the parameters.

I've seen this approach used to model how doctors weight different patient factors when deciding to prescribe an antidepressant. Interestingly, when you compare the regression weights of the variables with the importance that doctors think that they assign to them, the results can be quite different. Instead of weighing up the balance of factors, doctors tend to respond to the most obviously abnormal factor!

Here's an example : Article Using the method of judgement analysis to address variations...

I think the time it takes. A data of high weight is the one someone talks a lot about and the one with small weight is the one that someone doesn't want to answer.