Excel
Create Dataset with at least 30 observations
Create your own dataset that includes at least 30 observations for four different variables, i.e. minimum 30 rows, 4 columns
You have created a copy of the raw data and labeled the sheet ‘Raw Data’
You have renamed the sheet you are working with ‘Analysis.’ The first step is to add at least one column and transform an existing variable (e.g., from text to a quantitative variable). You can choose any variable you want to transform. Hint: you should most likely be using an if() function to complete this step.
You have provided descriptive (summary) statistics for each variable in a new sheet named ‘Descriptive Stats’ – refer back to the first weeks of class to see what is meant by descriptive statistics. Use your own judgment on how to best show the data.
You have described the distribution of at least one variable in a sheet named ‘Distribution’ – you should use a histogram and its chart to show the distribution of the variable. In cell A1, explain why you think the variable has its distribution. Are there any outliers? How do you know?
You have completed a t-test on comparing two variables in a sheet named ‘t-test’. In a cell of your choice, explain what the t-test tells you in plain English. The description should be at least 10 sentences. You need to describe a reason or multiple reasons why that test would make sense. It is ok to make some assumptions and be creative – in other words, tell a story about the data.
You have completed either a decision making (input/output) model based on the data (using solver, expected value, or a custom model similar to what we performed in class) OR a regression and explained the regression in a sheet named ‘regression.’ In a cell of your choice, explain the input/output model or what the regression tells you in plain English. Again, you need to describe why you chose this methodology to look at the data. It is ok to make some assumptions and be creative – in other words, tell a story about the data.
You have explained in a sheet named ‘Discussion’ describing what data is missing that would help make a better decision. What is the decision you could make from the data set if you could magically snap your fingers and have the data. What type of statistical test might you run to complete the analysis with the new data.
