We are officially out of school (unbelievable how time flies, right?)! The past few days, I have been enjoying getting more people to participate in my lip gloss blind consumption test and training for my new job at Starbucks. In case you haven't been following my research until this point, welcome to my blog! The goal of my research is to figure out more about consumer decisions between natural and unnatural makeup brands, specifically whether consumers choose between products primarily based off of differences they actually feel while consuming them or based off of societal influences pushing them to consume a certain type of product (e.g. health movements pushing consumers to purchase products made with all natural ingredients, as opposed to products containing unnatural ingredients).
Yesterday, I went to Skin Apeel Beauty Bar (a makeup store on Avenue of the Fountains in Fountain Hills) again and asked clients and customers if they wanted to participate in my lip gloss blind consumption test as they came through the store. I ended up getting 13 people, and now I have 34 of the 50 participants I need. I am going back again this upcoming Thursday, and there should be a lot of traffic going through the store that day, so I am confident that I will have all (or nearly all) my participants tested by the end of the week.
I have also inputted all my results into a Google Sheet I created. For each participant (assigned a number), it includes her blind consumption scores for each product, the 4 products listed from most favorite to least favorite (based on these blind consumption test scores), the 4 products listed from most favorite to least favorite (based on the exit survey, which asked participants to order the 4 lip gloss brands from "most favorite brand" to "least favorite brand"), a mismatch score (which indicates how "off" or different the blind consumption order of favorites differs from favorite brands list), and a natural product movement influence score (based off of the multiple choice questions in the exit survey meant to gauge how strongly the participant feels about consuming natural cosmetic products over products that contain unnatural ingredients).
Thus far, I am finding that almost all participants have similar mismatch scores regardless of how influenced they are by the natural product movement. I created a graph in excel plotting each participant's mismatch score (a number 0-8, with 8 indicating the most amount of difference between the blind consumption test and the participant's indicated preferences) with their natural product movement influence score (a number -20 to 20, with -20 being highly uninfluenced and 20 being highly influenced). My hypothesis was that participants highly influenced by the natural product movement would have more mismatches than those who had little influence. If this were correct, the graph would look more linear with a positive slope, indicating that the number of mismatches increased as the influence by the natural product movement increased. However, so far, it seems that the number of mismatches is relatively unrelated to natural product movement influence, as shown by the lack of trend in the scatter.
So, there you have my progress so far. Moving on to how I feel about being off on my own now with less guidance from Ms. Haag -- it would be a lie to say that I am completely confident in how I am analyzing my results and how I will write my results section. I am hoping that is one thing we can start to discuss in our meeting tomorrow so that I can have more direction on what will be most useful to include. Right now, the pitfalls I am most concerned with is unclearly presenting my data (for example, I have an uneasy feeling about that graph being hard to follow/interpret) and including too much/too little information for the sake of answering my question (e.g. should I add up/average all the participant blind consumption rankings for each lip gloss and include which one is most liked to determine which product was actually deemed the "highest quality"?).
(680)
Audrey -- your project is interesting in terms of what you're to do in the third trimester. While many of your classmates are busy trudging through their samples, your data collection sort of happens very quickly and in spurts, which, in theory, leaves you without much to do for these next couple of weeks. However, I think your instinct is absolutely right, that figuring out how to narrate the story that the data is telling you is going to be an arduous task. I can tell you right now that while the graph is likely very telling, I got lost in what you were trying to say about it. There are a lot of outcomes that your data could elicit, so we need to get started in thinking about 1) how to articulate the potentials and 2) how to analyze the data. Therefore, I anticipate that your next few weeks will be challenging, even though your data collection happens rather quickly.
ReplyDeleteHey Audrey!
ReplyDeleteI really like how much progress you've already made! Your project's methods seem well thought-out, especially with the three-pronged approach of the blinded test, the exit survey, and the test to measure the influence of the "natural" make-up movement. Since I'm joining your blog later in the year, I'm just curious about how you blind the lip gloss -- did you put them in identical containers since the brands have different shaped containers?
Also, what in your eyes makes a lip gloss better than another one? Why do people prefer one over another? Look? Feel? (I'm probably astounding you with my ignorance about make-up). I feel that talking to participants and asking them a few questions (like observations, so not a full qualitative study) may increase the robustness of your data, as you could potentially explain choices.
About the data -- don't be discouraged by the lack of correlation! Perhaps, rather than measuring number of mismatch, you could specify what brand the mismatch was associated with? This could provide some insight about what specific brands have large differences in preference and brand reputation. Also, what is your correlation coefficient right now?
Well, I got super into this comment... sorry for all the questions! I'm just fascinated by all this!
Cheers,
Yash
(212)
Hey Yash! No need to apologize! I am happy that you are fascinated by my research haha. I put multiple layers of black duct tape around each tube to the point where the containers looked virtually identical (the tape was so thick and the way it was stuck on morphed the shapes). I chose lip gloss primarily because it was economical (using different disposable lip gloss wands for each person and having them just dip them in the tube only required me to purchase one tube from each product), but also because it is pretty straightforward to apply (just put it on a wand and swipe it on) and because one color (a pink, in my case) could be used on different ethnicities and does not have to be "color matched" for each individual (like, foundation, for instance). I am having people rank the lip glosses on a hedonic scale 1-9 (1 being "bad" and 9 being "good") for the qualities of color (although I chose the most similar colors I could in all the brands, the pigment quality could be different), shine, taste, and texture because these are the qualities I found addressed in a lot of makeup reviews.
DeleteAlso, thanks for all your suggestions! Super helpful! I am scheduled to calculate the correlation coefficient after I get all my data, so I will include it in my next blog. :)
Hey Audrey!
ReplyDeleteYour organized and methodical way of completing your research is inspiring to say the least. I greatly appreciated how you put kind of a mini summary of what your research is about at the beginning for people not in your class like me. And in hindsight, I really wish I had done the same for you!
In terms of your data collection, I am really happy to hear that you will be done early because of how diligent your process is. When you were providing all the data you have gotten, I found it a little confusing to follow how each little conclusion for each part (the blind consumption score, the exit survey, the mismatch score, and the natural product movement influence score) all connected to make a final greater conclusion. It definitely makes sense if you think about it, and I am sure you will be wary and explain all the close links in your actual results sections, but I just thought I would point it out.
It is interesting that a correlation is not immediately eminent as of now with most of your participants. This can definitely still be a significant conclusion to draw, as maybe there is not a large effect of the feel of natural ingredient-make up on consumer choices, and it is more of a push from society/social justice movements. I know I personally buy organic brands not really for the quality, but more because I appreciate the message it represents. This study is so interesting because on a molecular level the makeup companies can tell us how organic should feel better or at least be better for us longterm, but I have never seen a study like yours that is judging brands just on face value, on what a consumer sees and feels immediately.
You research is so cool! And I think adding up the scores and finding a "highest quality" product is certainly not necessary, but could simplify your results for the reader and provide an interesting insight for sure. I felt like the graph was a little hard to follow, but I think with the explanation underneath and the tables to supplement it and refining from Mrs. Haag, I am sure there you will present it clear enough. That is definitely where you should direct your efforts once your collection is done.
Awesome work!
Hi Audrey,
ReplyDeleteI just checked the clock and realized its the first week of February. And hey, you're like 70% done with your data collection. As an afternoon student, I am not very familiar with your project, but your intro gave a great background and quick overview of your project. I should really be incorporating something like that into my highly technical blog.
In terms of what you found, don't be looking for the trend line. It is okay to have non conclusive results because the lack of correlation is also a separate conclusion in it of itself. When I first look at your graph, I am a bit confused. I don't see how the graph is designed. (This may be in part due to no trend lines). Maybe draw your expected hypothesis line on the graph to better illuminate the purpose of the graph. As it is right now, the graph is a bit confusing.
But it sounds like you have your work on lock like a cell block.
-
Ashwath V.