Overview
Make a presentation given a data set that covers the material covered In class.
For the analysis, the groups must use all the analysis we learned in class. For example (but not limited to), you have to provide summary statistics (and why you selected it), plot the data to discuss data distribution (whether skewed, or is normal, or outliers), and then conduct a hypothesis test to investigate your data. You also have to select what type of hypothesis test you will conduct (whether it's x_bar or p^hat, or group differences, etc).
4. The presentation will be designed for 6 minutes and must contain following:
a. What is your topic? An example would be, “Has the traffic congestion in LAX changed in the last 5 years?”. This slide will be called “Research Question.”
b. What is the problem? This slide is 3 to 4 lines on why you selected your topic. An example would be, “Many people have been complaining that the LAX's congestion has increased and is causing traffic jams. The traffic Jam in LAX creates pollution, causes loss of resources, and thus needs to be investigated”. It doesn’t have to be a lot of information. I need a 3-4 line justification on why you chose it. This slide will be called “Problem Statement”
c. What is your method? This slide will describe how you collected your data, what methods you will use to analyze it, and any problems with the data. For example, you can say, “We collected the data from bbc.com. We first conducted a summary statistics analysis to quickly explore the data. Then we used hypothesis testing to check our topic. We also used conditional probability to investigate the independence of different variables.”. Please note that this is an example; you must describe what you did in detail. Also, identify potential limitations of your data or methods. This slide will be called “Methods”
d. Now, you will present the results of your analysis. The results slide contains most of your grade points. In this section, I don’t only want the calculations. I want you to look at the numbers and explain the results. The better the explanation, the higher you will be graded. These slides will be called “Results and Discussion”. For example,
i. The results slide 1 will describe the results from your summary statistics. You will plot it to show any correlation or skewness and then tell what and why you picked a particular summary statistic (mean/median/mode & Range/IQR/SD, skewness and outliers). You will talk about the results.
ii. The 2-4 slides may describe your hypothesis testing or your normal distribution analysis
e. The last slide will be the conclusion, where you will answer the research question in one line.
- Few examples (but not limited to) of question I can ask during the presentation are as follows ( some but not limited to):
- How did you ensured a representative sampling for your study? (is your sample size enough or have you though out any obvious bias, etc.)
- What are the different type of variables in your data (categorical vs numerical, etc.)
- Which central tendency and dispersion metric did you select? and why? ( You may answer it by plotting a histogram and showing skewness. Also can talk about outliers)
- What is the difference between population distribution, sample distribution and sampling distribution?
- Which test are selecting for your hypothesis and why (t-test, z-test, t-test paired or t-test group differences)
- Did you check for CLT? Why do you need to check for CLT? what happens if I don't check for CLT? (You will use the concept of random error and sampling distribution to explain this)
- What CI did you select? and why ( think about type 1 or type 2 error and describe the risk)
- Be comfortable with the material and look out for conceptual understanding
- Potential Mistakes while making the presentation
- Using wrong central tendency and dispersion metrics. (DO NOT calculate all the metrics)
- Not plotting the graphs to show skewness and outliers
- Not investigating the data prior to analysis
- using t test, but using z score and z* to calculate CI
- confusing between z-test, t-test or ANOVA
- wrong interpretation of p value or not knowing what that is
- missing CI or wrong range
- Not interpreting the hypothesis test properly or missing CLT tests
- Not providing written information
- not providing any probability calculations
I have already chosen my data set I am doing the difference between male and female social media use and have attached my random sample below.