There is a large literature on research designs in mixed methods research that might be of use to you. The book by Creswell and Plano-Clark provides a good introduction to this field.

The most important question would be the use that you want to make of the quantitative data, and how that relates to what you want to do with the qualitative data. In mixed methods terms, this is a question of how you want to "integrate" the two sets of results, and thus the purpose served by each set of results .

I'll see if my university has that book.

Well here's the thing, the two sets are performing two different functions. The first set, which focuses on individuals with military backgrounds, is only there to establish general congruencies in personality and temperaments with non military individuals (which, if I'm going by my d scores seems to be a reasonable assumption to make). As far as the data seems to be concerned, the only significant effect sizes occur in Orderliness (in favor of Military) and Compassion (also in favor of military), which was surprising until I looked at the way I worded the relevant question in my BFAS. Other than that, the data collected for this group is just a side-effect of the entire study, and not necessarily important. My goal is to examine whether military experience acts as a orienting factor for video game behavior.

The second group, adolescent gamers, is there to provide a low-resolution picture of participants personality structures. I know a small BFAS is not comprehensive, but its simply supposed to (a) replicate the findings of Borguis et al.'s 2007 study about personality development across adolescence, so that I can (b) look for evidence of personality influence on moral intuitions during gameplay. The thesis here (in response to Joeckel et al.'s [2012] assertion that adolescent gamers seem to act randomly when presented with in-game moral dilemmas in order to facilitate game continuity), is that in the absence of an "experiential lens," players behavioral intuitions are motivated by their personality structure.

Now for the rub ... I'm expecting players, as a generality, to act pragmatically in the games (i.e. acting in whatever way produces the "best" end result), a form of moral disengagement they'll justify with comments like "it's just a game." My prediction is that I should see evidence of experience or personality in the qualitative response e.g. "I did action X because it made sense in the game, but if it were real I would have done [Y]."

I know that was waaaaaay more than you probably wanted know, but I just wanted to explain why the quantitative data is not necessarily the most important thing to my research. To the best my knowledge Cohen (1988) makes it seem like a d value for comparative between the groups, and a Pearson correlation between the traits and the two specific behavioral patterns seems to be enough.

Hi Ryan - I like your 'up front' confession. Your further comprehensive response to the ever-reliable David - is a little concerning. I can see why you might be confused. You seem to 'mixing your mixed methods'.

If you want to promote this as a quality mixed methods study - you can't approach it from the point of view that the quant phase (no matter how lower case) is merely discussion or just a 'side-effect'. It has to be a planned and integrated component.

You also seem to be 'quantifying' your' qualitative data. You are predicting certain responses - and that is not the place of qualitative research. It is to make sense of what evolves form the collected narrative.

Your latter comments on Cohen and correlation don't seem 'to correlate'. You've stated that the quant outcomes done't really figure.

Dean Whitehead ,

In retrospect I realize I did a really poor job of explaining how I planned on utilizing the quan data in this study. To start, I've been kicking around converting the d scores I got when comparing the "treatment" and control groups of the study (technically the control group in study two is the treatment group, since "age" is the treatment) to Hedge's g because of the small sample size. I know that doesn't address the concern you've discussed, but I wanted to keep people abreast of my changes.

So for the military group (n=31) my hypothesis is that military experience creates an inherently conservative lens through which individuals see the world, so the primary goal for the quan data in that regard was to establish a relationship between the two groups and look for evidence of differences in key trait domains /aspects linked with conservatism, which seems to be the case. A significant effect size in the direction of military people were indicated in conscientiousness (0.26), most prominently in aspect Orderliness (1.22)> I haven't recalculated the data to g but I can't imagine it is going to be wildly different. The other significant finding was an effect in the direction of the nonmilitary control group in Openness (.2), specifically creativity (.28). From the research I've read those seem to be two areas specifically related to conservatism. That was basically all I intended to do with that data: look for evidence of an "experiential lens" then hopefully find the markers of that lens (using Moral Foundations Theory as a guide) in the QUAL responses.

As far as the adolescent group (age 13-18) is concerned, I wanted to (a) measure personality for each of the respondents (n=13, control group same size, mean age = 26 ) so that I could look to see if their personality influenced their in-game behavior in the absence of a fully formed moral lens (related to Joeckel et al.'s findings), but in order to do that I felt as though the data needed to be at least remotely similar to other findings on the same demographic, which for the most part I did (with the exception of agreeableness--for some reason that indicated a .3 effect in the direction of adolescents--I'm chalking that up to the relative lack of sensitivity of my instrument i.e. it's only 10 items, and not longitudinal). Furthermore, I know that traits in that age are essentially in a state of flux, which further problematizes the small sample size. An unexpected benefit of this part of the study is that almost none of the adolescents indicated strong religious values or political stances with the exception of two, so the plan is to look for indicators of personality structure as an orienting factor in the QUAL responses, as opposed to a moral lens formed and contextualized by life experience.

So the goal in the end is taking this quan data and saying "Here're these markers that are consistent with other studies about moral lenses (experience-military), and personality development (no experience-adolescents). Do they manifest in the expressed moral intuitions of players during gameplay?"

Last bit: the correlation section is just a little ancillary thing I included in the study. For example, in Metal Gear Solid V, enemies can be injured (and subsequently bleed out) during combat, and players have the option of "rescuing" them by airlifting them to "Mother Base." I was just curious to see if that behavior correlated with any of the personality measurements.

Ryan - that makes more sense. You seem to be hiding your light under a bushel. What you identify here sounds useful and interesting - and 'is what it is'. You may not have significance in some areas - but that's not a problem. To me, the correlation is actually one of the more important areas and might be your main point of originality. I would pursue it more than 'ancillary'. Back to your qualitative phase. You still have to move beyond the 'looking for specifics'. If what you would 'like' to see emerges naturally from the collected narrative/data - then all well and good - but don't give an impression to markers/reviewers that you have 'cherry-picked'.