Study design and data analysis

#1
Hello everybody, I am doing quantitative research for the very first time and am currently a bit lost.

My research deals with the experiential advantage: People derive more happiness from experiences than they do from material goods.
This is due to 3 reasons:
1) Experiences are more self-expressive
2) Experiences are more social
3) Experiences are harder to compare (for example against better options etc.)

Interestingly I found out that customization provides material goods with those very same qualities: When you customize a material good (e.g., design a sneaker yourself), it becomes more self-expressive, more social and harder to compare. Even though the effects of customization on experiences are currently unknown, my hypothesis is that people will derive a similar amount of happiness from a customized material good and from a customized experience. My reasoning is that customization adds “experiential” qualities to a material good, thereby leaving you with a purchases combining characteristics from both the material and experiential world. In contrast, if customization has the same effect on experiences, I would just be adding more of the same things.

I hope this was somewhat clear. In any case I am currently working on my study and this is where I’m having trouble. Basically I’d like to do a 2x2 between-subjects design with happiness as the DV, and customization (customized vs. non customized) as well as purchase type (material vs. experiential) as the IV’s. My plan was to then do a two-way ANOVA for my hypothesis (people derive a similar amount of happiness from a customized material good and from a customized experience).

This probably is a very stupid question, but how could I include those three qualities (self-expressiveness, socialness and comparability of the purchase) into my research? Basically I'm hypothesizing that customization makes material and experiential purchases more self-expressive, social and harder to compare, but that this effect will be stronger for material purchases, thus evening out the experiential advantage seen in non-customized purchases. Would it be possible to conduct a mediation analysis after the ANOVA to figure this out? What would be the easiest way to accomplish this?

Any kind of help would be greatly appreciated
 

Karabiner

TS Contributor
#2
When you customize a material good (e.g., design a sneaker yourself), it becomes more self-expressive, more social and harder to compare.
Really the resulting good? Or rather the active process of customization?

My reasoning is that customization adds “experiential” qualities to a material good, thereby leaving you with a purchases combining characteristics from both the material and experiential world. In contrast, if customization has the same effect on experiences, I would just be adding more of the same things.
I must admit that I do not fully understand what you are saying here.
I hope this was somewhat clear. In any case I am currently working on my study and this is where I’m having trouble. Basically I’d like to do a 2x2 between-subjects design with happiness as the DV, and customization (customized vs. non customized) as well as purchase type (material vs. experiential) as the IV’s.
What exactely does " customization (customized vs. non customized) and purchase type
(material vs. experiential)" mean, what will be done in the experiment?

Apart from that, you might perhaps think about an "mixed design" with between-group and
repeated-measures factors. This way, subjects serve as their own control, and this reduces
statistical error and increases power.

What would be the easiest way to accomplish this?
Perhaps structural equation modeling.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891.
Vanderweele, T. J., & Vansteelandt, S. (2014). Mediation analysis with multiple mediators. Epidemiologic Methods, 2, 95–115.

With kind regards

Karabiner
 
#3
Hi there, and thank you very much for the prompt response.

Sorry for being a bit unclear, I tried to keep myself as short as possible. Without customization, people tend to derive more happiness from experiences cause they are more self-expressive, more social and harder to compare against better options. These are the "experiential" qualities and they have all been found to increase happiness (albeit more so for experiential purchases). I'm trying to figure out if this experiential advantage remains, once customization is thrown into the mix.

There are studies saying that customization equips material goods with the very same qualities that are responsible for the experiential advantage seen in non-customized purchases: For example you can make a shirt "more you" (more self-expressive) by designing it yourself. Furthermore, just like you said, the customization process itself is an experience. Thus, after customizing a material good, I am left with a memorable experience as well as a tangible material good that now has the very same qualities that are usually seen in experiences. So far there is no research on the customization of experiences, so I don't know if they will become even more self-expressive, even more social and even less comparable. (That's one of the things I'd like to find out in my study) However, even if this were the case, I would just be adding more of the same experiential qualities that it already has in abundance to begin with. And since those experiential qualities were found to influence happiness, I'm hypothesizing that customization will increase happiness for both experiences and goods, but that this increase will be stronger for material goods. (cause I'm adding new qualities, not just more of the same)

Regarding my experiment, I would like to have four conditions: experiential/standard, experiental/customized, material/standard and material/customized. I built two websites, one offering sportswear and one offering sports classes (e.g., a swimming class). Depending on the condition, the participant should either pick out a standard sports class/shirt or customize the sports class/shirt to their liking. Afterwards they should fill out the study indicating how happy this purchase would make them and how self-expressive, social and comparable they find it to be.

Basically I'd like to indicate the happiness levels of the four conditions, but also how this happiness came about. So if my hypothesis would hold true, I'd like to be able to say that yes, people derive a similar amount of happiness from customized goods and customized experiences, and this is because the positive effects of customization (making a purchase more social, more self-expressive and less comparable) are stronger for material goods. Or, if I'm wrong, I'd like to be able to say that customization does not actually make a material good more self-expressive and so on.

I didn't have a mixed design in mind because I'd like to keep everything between-subjects, but I've looked into the SEM recommendation and I think it fits what I'd like to do. Thank you very much for this.

In the mean time I have also thought about doing a MANOVA with four IV's (happiness, self-expression, socialness, comparability). Do you think that would also be a possibility?
 

noetsi

Fortran must die
#4
Structural equation models are not for the faint of heart. After a year of studying them in a graduate program - I concluded they were complex. If you have not worked with it before and don't have a very strong statistical background I would be cautious about using that.
 

Karabiner

TS Contributor
#5
Structural equation models are not for the faint of heart. After a year of studying them in a graduate program - I concluded they were complex. If you have not worked with it before and don't have a very strong statistical background I would be cautious about using that.
Noetsi is right. Maybe Andrew Hayes has some alternatve to offer http://afhayes.com/spss-sas-and-r-macros-and-code.html

What do you mean by "a MANOVA with four IV's (happiness, self-expression, socialness, comparability). " ? Do you mean 4 dependent variables? That willl not answer yoiur research questions, as far as I can see.

With kind regards

Karabiner