Assignment Checklist for Data Analysis

Overview

This page is not intended to make your life harder and more "tasky". Rather, the objective here is to help you remember key things that you need to know for your assignments. In the first one or two assignments, I will typically remind you of specific things you need to do. However, beyond that point, I will expect you to remember to adhere to this way of doing things. Some of these requirements are not simply good practice for this class, but rather, are good habits for doing data analysis in general.

You are not, of course, required to meet objectives until we discuss them in lecture. For example, if we have not yet discussed how to create a scatterplot, then I certainly don't expect you to worry about the specific requirements of scatterplots just yet!

I strongly recommend that you refer to this document before submitting any assignment. It's a great way to make sure you don't miss any obvious / easy points.

The Checklist

  1. Submit as Google or Word Document: Your submissions should always be in DOCX (Word) or Google Doc format. Do not submit as PDF.

  2. Always show your work: As mentioned in the past, I don't expect you to go into painful detail of every step, but the grader must have a decent idea of how you arrived at your answer.

  3. Describing a distribution: When asked to describe a distribution, this should include at least one chart (typically a histogram or a boxplot -- and including both is okay), and descriptive statistics (center, spread, 5-number summary).

  4. All graphs should be included in your document.

  5. Include title and labels on graphs: When drawing a graph, you should always include a title, and at the label on at least one axis (histograms and boxplots), and for some graphs, such as scatterplots, you should have labels on both the x and y axes.

  6. Create vectors in R: When you are provided with a series of values and expected to do things with them using R, you will almost certainly want to place those values inside an R vector.

  7. Always include your R code in your assignments: In addition to providing numeric solutions from your R output, you should, of course, also include any R code you wrote to generate that output.

  8. Be prepared for questions that can not be answered: I will be throwing these at you from time to time. I promise that my goal here is not to trick you! However, you must be aware that in the real world, people will often come up with statistical and mathematical conclusions where they are not appropriate. You will come to see that I am far more interested in pushing all of you to ensure that you grasp concepts, than I am in making you spend hours crunching numbers. Types of questions that can not be answered:

    1. Not enough information: As a very simple example, if I gave you a mean but did not give you an standard deviation, then you would not be able to answer the question because you have not been given enough information.
    2. Inappropriate for the type of calculation being asked: If, for example, you are asked to do normal distribution calculations, but you plot the data (and you should always plot the data!) and it turns out to be skewed, then doing calculations that require a normal distribution is inappropriate and would lead to faulty conclusions. Another example would be if you were asked to generate a linear regression model - but you plot the data and see the relationship is not linear!
    3. In cases like the above, all you have to do is to indicate as much your answer and you will receive full credit for that problem!

     

Other Thoughts

Real world application: As an educated citizen of the world, it is important to be able to interpret statistical scenarios without necessarily having expertise in the particular area. At some points in this and other assignments, you may be asked to consider some real-world ramifications of the calculations you are being asked to determine. Do not take these questions lightly as they may have right and wrong answers. The key is to spend at least a little bit of time thinking about them.