For those of you who have been reading Statistical Sage for a while, you know … we’re all about helping students mastering the underlying concepts in statistics. With that conceptual understanding, students can then begin to use statistics as the tools that they are, for answering important questions. However, if you are missing critical concepts (e.g., what is a sampling error, and how do we find it), then you may miss a critical step (e.g., like assuring that your data has been selected independently, otherwise your estimation of the standard error will be as useful as flipping a coin.) Yes, at Statistical Sage we focus on helping students develop a deep understanding of statistical concepts. However, when I read the article in the Teaching of Psychology http://top.sagepub.com/content/38/4/247.abstract?rss=1 Staldler and Olson (2011)’s article t for Two: Using Mnemonics to teach statistics, this gave me a bit of a pause.
Are we, in this era of immediate gratification, doing our students a disservice by telling them that memorizing a few “trick” phrases can help them master statistics? Yet, it is true. When I teach about the independent and related t-test, I sing the song … “Tea for Two,” I also sing the song “‘Tis a Gift to be Simple” when I review the canon of science, parsimony. I even use hand signals and sign language when I am teaching. Typically students don’t even recognize my use of American Sign Language when teaching. Yet, when they get confused on the test and ask questions, often just a repeated hand symbol will be enough for the students to access the information… not just the word, but the entire concept. And, yes, I’ve even hummed a few bars of a song I sang in class and it often helps student retrieve all of the information they need. Thus, as I thought about mnemonic cues, I thought that they are just a way to ease the trace to the stored information. If the information was never stored, it’s not going to get the students to master the concepts.
Thus, I encourage everyone to help students with mnemonic cues, but don’t be fooled into thinking that mnemonic cues, alone, will be enough to help students master and apply important statistical concepts.
Now, I will share my two favorite mnemonic cues that I use for Error.
As you know, there are two classes of error in research … observational error and decision making error.
Let’s take observational error first … let’s face it, error is a MES (with one S). We have Measurement Error, Experimenter Error, and Sampling Error. We use Statistics to estimate the Sampling Error, that is error due to individual difference of the subjects who just happen to be in our sample. I find that by using this basic mnemonic, students are more apt to recall all three types of observational error.
There is, however, a second type of error … that is decision making error. Now, these tremendously nondescript labels for decision error (Type I and Type II) are easily confused. This simple trick helps students … Type I error is a False Alarm (the probability of finding a statistically significant difference when one really isn’t there), while a Type II error is a MiSS (The probability of missing a statistically significant difference when one was really there). Notice that Miss has two S’s for Type TWO.
Sure, I do several activities to help students to understand the concepts of sampling error and Type I and Type II error. In truth, the concepts are a bit easier than matching the names to the concepts, thus the mnemonic device helps students retrieve the information.
I highly encourage everyone to enter into the use of mnemonic cues carefully. Make sure students understand the underlying concepts before teaching them the mnemonic trick, and it will do what is it suppose to do, help lead students to the information so carefully stored in their brain.
If you have mnemonic tricks you use in your stats classes, I encourage you to post them here.