First, here is a brief synopsis of what I ended up doing last week:
Monday: I started analyzing the results of my data and found that performing confidence intervals on my data takes a bit longer than expect. To speed up this process, I plan to create an excel sheet that will run the numbers quickly.
Tuesday: I met with Mr. Peacher and thankfully my idea to use a confidence interval on the slope of my season versus percent graphs was correct.
Wednesday: More finding the averages of my data within the seasons.
Thursday: I edited my literature review and methods and continued to find the averages of my data within the seasons.
Friday: I met with Mrs. Haag and again kept finding the averages (if you cannot tell this is what is taking the majority of my time to complete)
Plan of Attack for Data Collection:
I am almost done with finding the averages in all the seasons for every single piece of data I have collected. On Tuesday, I plan to finish this off and begin graphing my data and finding confidence intervals for the trend lines. Additionally over the long weekend (Sunday and Monday) I will be editing my literature review on the airplane so no time goes to waste.
Here is how I plan to analyze all of my data. First, I put the data set into excel to create a basic scatter graph. Next I find the line of best fit for trend lines. After that I preform a 95% confidence interval on the trend line's slope since I used a random sample in my research. Over Tuesday and Wednesday, I am going to try to run all of these graphs.
The hardest part of my research is going to be how I end up organizing all of my results from all the graphs. How do I weight the sections? Does one section matter more than the others? What if nothing really happened throughout the years? I know I am going to have to organize these results multiple times to see which way makes the most logical sense and best represents my findings, especially since the paper can only be 5000 words I am going to need to be very mindful in what results I prioritize. I plan to look into some psychology journals to determine if consumers are more drawn to competition or drama to see if any of these findings will be helpful in organizing my data. This is the part I am the most unclear about and therefore it deserves the most time. Additionally, the confidence intervals and graphing will not take that long as I have created an excel sheet that calculates the numbers rather quickly.
Word Count: 444
Catherine -- is there anything in your literature review that speaks to which elements are prioritized? That might be a helpful place to find justifications for considering and weighing the different elements against each other.
ReplyDeleteAlso, if nothing happened throughout the years, I want you to think about what you saw while watching all of those episodes and see what you intuitively recognized. Thereafter, you can see if the data aligns with any of your intuitions (even if it's not necessarily about it being more formulaic). I really want you to mess around with what you could extract from the data, so when we meet, we can start to outline the results section together.
Hey Catherine!
ReplyDeleteIt's exciting that you finished coding all of the seasons of Survivor... kudos to you for putting in the time to finish that! As a new-comer to your blog and someone not familiar with your project, I wasn't sure what your coding data looked like, but what I gathered from your post is that you have been assigning some sort of point-value for specific categories/markers (along with written observations, of course) and now you are plotting the relative point values for different seasons. Is this correct?
If so, what is your scatterplot's independent variable? Is it season number? I'd probably like a slightly better explanation of what your scatterplots show. Also, you could potentially find the correlation between TV viewers per season and level of competition and drama.
Could you even potentially use some t-tests to find some p-values to see if the level of certain categories was more than expected? Just some thoughts...
Finally, did you end up using the Cohen's kappa coefficient to check for intra-reader reliability? That is an important way to prove reproducibility, which is crucial for your study.
Sorry that I'm a bit unfamiliar with your project, but hope this helps!
Yash
(200)
I really like the idea of you including some sort of analysis that looks at the prevalence of certain elements as related to ratings.
DeleteHey homegirl!
ReplyDeleteIt sounds like you're staying on track with the schedule you created and I'm glad you decided to create a spreadsheet to make your life easier so early in the game. Also, I think meeting with Mr. Peacher and coming up with a solid plan to do the statistics behind your results was a really great thing to do because you have an idea of what your results section should look like.
Also, like Yash, I'm not really to familiar with your project. So from what I understand, your analyzing all of the seasons of Survivor, but I don't know what you're looking for. Are you looking for whether the seasons and the participants' actions become somewhat formulaic? If that's the case, what does your sheet look like that your plotting the differences and similarities between the seasons? If there are variables you're trying to find the relationship between, you could do a Chi-square test to first determine if there is some sort of a connection and then find the correlation between the two.
Overall, it seems like you have things under control!
Hey Catherine! So I just wanted to begin by saying that I love the background. I think it is absolutely amazing. Also, I heard you were interviewing this past week while also conducting a ton of research. Kudos to you for managing your time so effectively and efficiently. Also, I think that meeting with Mr. Peacher was an excellent idea to ensure that you correctly use statistics to get the most out of the data that you collected. Keep on working hard and I am excited what kind of results you can get from your data! Also, by the way, did you use a statistical program to analyze your results? If so, could you let me know what it is called?
ReplyDeleteThanks,
Brian