This past week has been extremely important for us in AP Research. We have finished up the data collection and now our projects are really coming into shape. Specifically with my project I had 40 total participants, which is exactly the number I was looking to get, over the course of four days. As the experiment progressed, it was really exciting to see the data come together and begin to come to some preliminary conclusions based on my observations. Right now, I have data that records whether each participant hit or stay on a 15 after exposure the the effect for their group as well as observations of their behavior and data on any other ambiguous hands the participant saw. In order to qualify my data and better understand it, I had each participant fill out a survey about their inherent riskiness levels. Now that the data collection has been finished, I have all the clues to the mystery of which effect has the largest influence on blackjack behavior. Just like Sherlock, I need to find the best way to put the information together.
In order to understand which one is the most influential effect, I need to understand which effect increases people's risky behavior the most. In order to measure risky behavior, I recorded whether each participant hit or stay on a 15, which is shown to be indicative of risk taking in blackjack. The risk assessment surveys that the participants took were also scored to better understand the inherent risk levels of each group. The risk assessment survey consisted of questions in which participants rated their likelihood of performing many risky activities on a 1 to 10 scale. I used a weighted average, with some activities being more influential than others, to give each person a riskiness score. I have calculated each person's riskiness score and taken the averages of each group. In order to understand how the effects change riskiness, I will compare the average score of each group to the percentage of times that people hit for that effect. Then I will try and calculate a change in riskiness to show how each effect is actually changing risky behavior. I have talked to Mr. Peacher and am planning on talking to him early next week about the data analysis to make sure it is sound. Right now, I have found that the control group hit 6 out of 10 times and the other groups have increased the amounts of hitting, with the near-miss group having all 10 people hit. Once I understand the inherent riskiness of each group, I can make conclusions from that data, but right now it looks very promising. This week while I'm trying to figure out my data analysis I may be like this.
But once the data has been analyzed, I will be in a great position and will be feeling more like this.
This week is crucial for the strength of my project, but I'm confident that it will turn out great and yield very interesting results. The fourth series of Sherlock came out recently and I'm so excited to watch it as well as solve the mysteries of my project. (533)
I think it's going to be very important to make the riskiness survey tightly converse with the amount of times groups hit.
ReplyDeleteMoreover, have you calculated the percentage of time that each participant hit? I know that some groups simply had more opportunities to hit than others because of the way the deck happened to be shuffled. Thus, that might be the most honest way to convey that data.
Max,
ReplyDeleteFirst of all, thanks for reminding me that Season 4 of Sherlock is now out (hopefully they decide to make a season 5).
I read through some of your previous blog posts to get an idea of where you were. You were having some troubles getting participants last week, so I am glad to hear that you were able to gather up your original goal of 40 participants. It seems that you have a lot of data to analyze. The surveys along with the number of hits and other data you collected comes out to be a lot of information. Do you currently have a system where you can effectively organize your data? If not, I think the best method for you right now is to make sure that you can organize your data before you start analyzing it, because organization is half of the data analysis—it can either hurt or help you throughout this week.
Overall, I think you are in a great spot and with a little bit more work, you will be able to get your results. Don’t give up, you’re almost there!
Hey Max,
ReplyDeleteIt is awesome that you were able to reach your goal and didn't have to sacrifice having less participants than you anticipated! I, too, almost ran into that problem and was freaking out. Now that you have all this data, I agree with Gursajan that if you have a clear system of organizing it, it can only help you. Especially when you are writing your results section, it'll be easier to present the information. I know my main problem is my data is kind of messy in a spreadsheet, so I will be trying to make it clearer to present in my paper.
In terms of data analysis, I think meeting with Mr. Peacher is a great idea to make sure you are on the right track. I agree with Mrs. Haag also in making sure your risk assessment survey is well connected with the data from amount of times groups hit, so that everything is cohesive in your results section leading to meaningful conclusions. How did you determine some activities were weighted more riskier than others?
Overall, you seem to be right on track for data analysis! Good luck with finding conclusions!
Hey Max! I think I have somewhat of a decent idea as to what you're doing since I was able to help you out by dealing the cards for your four trials. Because of this, I know that you had everyone in the near-miss hit and had kinda shaky results in the priming (which, given our resources, was really difficult to do, also do you think LeBron is still there?). Anyways, as to connecting the survey with results, how are you going to look at the data if for some reason all of the near-miss people were the most risk taking people out of the 40? Is there a formula or statistical method you will use? I know that you're gonna talk to Mr. Peacher but I was wondering if you had any ideas in mind, perhaps from some of the studies you based yours off on? Once again, I think your project is really fun and interesting, and I cant wait to see how you analyze your data!
ReplyDelete