Minimal Statistics Baseline
Minimal Statistics Baseline
In Module 1, we discussed how statistics is like a language. For the last seven weeks, you have been immersed in statistics and, hopefully, you have noticed that in this time you have become less afraid of statistics, and less nervous about your ability to work with statistics. Soon, you will be on your own and in order to maintain (and improve) your level of comfort with statistics, and you will have to have a plan. You’ve heard the saying, ”The person who fails to plan, plans to fail.”
With that in mind, please create and describe your “Minimal Statistics Baseline” or MSB. An MSB is the minimum amount of statistics you will commit to use going forward. The goal here is to set a minimum level that you know you will do consistently and not to aim too high. Rather than having a goal that is out of reach and that you may or may not accomplish, it is better to do something that you know you will be able to do consistently. It is always possible to do more if you feel like it. In your post, please describe your MSB and make sure you discuss the following:
- How often will you focus on statistics (e.g., daily, weekly, or monthly)?
- What kinds of activities will you do? (e.g., read a newspaper article and focus on understanding the statistics, look up what you don’t know instead of just passing over it; read a research article from the library; ask a question of someone who is an expert in statistics.)
- Why do you think your MSB is appropriate for your current level of comfort with statistics and where you choose to go with it?
There is no right or wrong answer; your MSB depends entirely on where you are on your journey to learn statistics. Your MSB can be as simple or as complex as you choose to make it.
Putting It All Together
The various statistical tests that you have learned so far were introduced to you as discrete topics, separate from each other. While it is important to study each test or technique independent of the others so that you can really understand and learn how to apply it; in practice it is rare that you will use just one test. More frequently, you will apply the tests sequentially and interpret the results to get a more complete picture of the data. Results that support each other help you to be confident of your findings. Results that seemingly contradict each other provide even more information they tell you the extent to which you can trust your results and also provide avenues for further exploration. For instance, you may find a significant correlation between two variables; but the relationship between the variables is not significant when you run a multiple regression with an additional variable. The outcome may indicate some relationship between the variables in the correlation and the ones you add in the regression.
When you are given a set of numbers to analyze, you will be given a general idea of what you want to know but not the specific test. So how do you proceed? You proceed in the way this course was taught. You begin by getting to know your data the types of data, levels of measurement, what each number means, the patterns, the average value, and the spread essentially what we learned under descriptive statistics.
Once you have completed descriptive analysis, you move onto inferential statistics. You test if there are differences between groups, associations, and regressions. Figure 10.1 on Page 220 of the Salkind textbook illustrates the questions to ask to help you decide with test is appropriate for your data and for the question you are trying to answer.
It is tempting to jump straight to the final test and not worry about descriptive statistics and other intermediate steps. However, statistical tests are only as effective as the person running and interpreting them. If you do not understand your data, it is difficult to run the appropriate test or interpret the results correctly. All the different tests we have studied provide varying views of your data and allow you to make more complete and informed assessments of the information you get from the data.
As Mark Twain said: There are three kinds of lies: lies, damned lies and statistics. Understanding your data and the appropriate test to use will prevent you from creating statistics that are lies and will also help you detect when others use of statistics is not correct.
Moving Forward
In our world today, we rely heavily on data and information, and it is very difficult to avoid the use of numbers and statistics. This course has given you some tools to understand the statistics that you meet in daily life and, we hope, has convinced you that understanding statistics will help you in your professional and personal lives.
In Module 1, we discussed that learning how to use statistics is like learning a language. Just as you wouldn’t expect to remember a language if you didn’t use it at all, you cannot expect to remember statistics if you do not use it all after this class. You do not need to remember every detail you have learned, but you do need to be familiar enough with what you have learned, so that when you need to know more, you can easily go to a resource and find out more information. Going back to our language analogy, you do not need to know everything you learned in the French (or Russian, or Mandarin, or Hindi) class, but you do need to know where your dictionary is, what the alphabet is, and how to look up the grammar rules. If you make it a point to regularly speak and read some French, and keep your familiarity with your reference tools, when you need more intensive French, you will be able to easily dive back in. Similarly with statistics, remaining familiar with the topics and the tools you use to learn more will help you dive back into the details when the situation calls for it.
