Your idea in this commence is fully pointed to the "Exploration" in a research process. In fact if someone wanna explore some concepts like as new job performance it is needed to apply an unstructured questionnaire/interview to capture and predict somethings more in his/her study.

Dear Amir,

Thank you for responding to my question. In fact, my question is different...I am not asking about research interviews but about selection interviews that are used to interview candidates who apply for a position....Most empirical studies show that structured selection interviews predict future job performance better than unstructured selection interviews.

Have a nice day

Sébastien

Hi Sebastien

As I have experienced in my researches and even executive works, you also confirm that if we make the unstructured selection interview not also is randomly examined but also let to explore future job performance too. Kindly please note that you make it open for receiving and achievement of something else which you haven't expected to find and predict the future. With my warmest regards

Sebastien,

Check this review https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2853669 by Schmidt. It has references for the latest most solid research on the topic, and it will also provide incremental validity analyses for 30+ selection techniques against GMA

Hope it helps

Thank you Gonzalo,

This article is very interesting and helpful but I do not find mention of studies about the incremental validity of unstructured interviews above and over structured interviews.

As you mention it, it is very detailed about the incremental validity of unstructured interviews above GMA.

It really makes me think that there is an important theoretical and empirical gap in the literature.

From table 1, it seems that both techniques have the same predictive validity on performance, although structured ones have add more validity to GMA than unstructured. I guess those figures say something about both techniques when compared to each other and when used with GMA. Considering that Schmidt applied a new range restriction technique for the analysis of old and new meta-analyses, I think the information is probably as solid as we have for now. I would say that regarding your original question, they both predict about the same but structured are better when used with GMA. Also there is the argument about reducing biases when using a structured format versus unstructured. This book from 2010 can provide more information on the topic of interviews: 

https://books.google.co.uk/books/about/The_Psychology_of_Personnel_Selection.html?id=TsxRkXKJ97EC&printsec=frontcover&source=kp_read_button&redir_esc=y#v=onepage&q&f=false

About personality, conscientiousness which has been the most reliable trait for performance, has lost some of its predictive validity when compared to the same Schmidt paper from 1998, due to the rage restriction mentioned before. The debate of personality and performance that has been going for the past 10 years seems to be leaning towards a hard to defend position for its use when it comes to performance. When it comes to other outcomes, such as leadership effectiveness, CWB or OCB, then personality seems to be a better predictor than GMA.

If you are interested, here is a link to a book chapter regarding personality research on selection practices:

https://www.researchgate.net/publication/309725988_Using_Personality_Questionnaires_for_Selection

This is a topic that I'm very passionate about and I would be happy to provide more information if you need. Hope this helps!

Chapter Using Personality Questionnaires for Selection

Thank you for the chapter and for your response. Yes it helps.

It is interesting that you mention personality because I think that the added value (incremental validity) of unstructured interviews could be due to the fact that they can measure personality (which is not the case of structured interviews).

Based on the following aspects: 1) the high value of unstructured interviews (as reconsidered by Schmidt recently), 2) the fact that unstructured interviews have good construct validity to measure personality and 3) that other ratings of personality seem to outperform self-ratings to predict job performance, I start thinking that unstructured interviews could maybe provide an important source of information about the applicant's personality.

Unstructured interviews have been maybe seriously underestimated by scientists. It seems logical because unstructured interviews have lower predictive validity than structured interviews but in the past we thought that they measured the same constructs. Now, I start thinking that competences are better measured with structured interviews and personality is (or can be) better measured with...unstructured interviews...To test that, I need to see if unstructured interviews have incremental validity above structured interviews.

If you find in the future some data about this question, could you please share them with me?

Thank you very much, Gonzalo

This is an interesting question. I think it has not been studied much because organisations tend to favour either type (high vs. low) structure, as opposed to considering using them in a supplementary fashion. If you wish to study this issue, it would be important to ensure that you use different interviewers to conduct either type of interview (to avoid carry-over effects between the types, etc.). But it is most likely to be an interesting tweak as the current assumption is that people like to use either of the two interview types (i.e., they are mutually exclusive) but perhaps you could use them together (e.g., to measure different things, as you suggest, or to predict different types of criteria such as task or contextual performance). Good luck!

Dear François,

Thank you for your reply and interesting comments and suggestions. It was first a curiosity because I do not intend to test this hypothesis (or at least not yet) but I was wondering if maybe it had already been tested...probably not yet (or it has been tested but with another purpose in mind, which is why it is so difficult to retrieve this information)..I think more and more often at the possibility that unstructured interviews have been underestimated by academics (and overestimated by practitioners). I am wondering if it is not worth investing time to reflect even further on this issue and to develop structured "unstructured interviews"... I know it is a strange expression but it expresses my idea that there is important information elicited in unstructured interviews that is not captured in classic "structured interviews" such as applicant personality...It is unfortunate that this information is not captured because it predicts important outcomes at work...Now the question is "how can we structure unstructured interviews"?...It will be maybe my focus for the next years