SCI Home | SCI Faculty Page | Cover Sheet | Reference List | Users Notes
Psychology
Cases - Case Notes
SCI Home | SCI Faculty Page | Cover Sheet | Reference List | Users Notes
Author - Keegan Greenier - Summer 1998
Background -
Teaching Notes -
Part 1—Concrete vs Abstract Memory
The purpose of this exercise is to demonstrate how one would study a psychological question scientifically. Suppose that we want to test if people’s memories are better for concrete words (things you can easily picture mentally) than for abstract words (words that don’t readily lend themselves to mental imagery). How can we answer this question scientifically?
First, we define our independent and dependent variables. Our independent variable (IV) is word type (concrete vs abstract). The dependent variable (DV) is the number of words recalled.
[If you want to get into the nature of hypothesis testing, the null hypothesis is that there will be no difference in memory as a function of word type; our alternative hypothesis is that memory will be better for the concrete words than for the abstract words. Naturally, we will either reject or fail to reject the null hypothesis.]
Lists of concrete and abstract words, along with brief instructions to the students are available in the following formats (WORD97, WORD95, MacWORD5.1, WordPerfect for Windows 5.x, Rich Text Format). After making photocopies, cut the page and give one half of the class one list, and half the other. (You may want to have students take out a pen and a separate piece of paper before you distribute the word lists.) Have the students keep the lists face down until you give the signal to turn them over. Instruct them that at no time during this exercise should they try to read through the back of the paper to see what the words are, as this will defeat the purpose of getting valid results. After giving the signal to start, give everyone the same amount of time to study the list (e.g., 20 seconds). When the time is up, have them turn the paper over, but tell them not to write anything yet. Wait for 30-45 seconds and give them the signal to write down as many of the words as they can recall, using their separate piece of paper. When you give the time’s-up signal, make sure they all immediately stop writing. Collect the lists and score each person’s list, or you can have them exchange papers and score each other’s (DV = number of words correctly recalled). Depending on time constraints, you could either just compare the means or, ideally, do a t-test on the results. A t-test is do-able in class even with just a calculator.
It is vital that students be randomly assigned to conditions. Randomly split the class in half to determine who gets which list. Random assignment helps to ensure that the only difference between our experimental conditions is the independent variable. Individual differences in memory ability, intelligence, sleepiness, etc., should be equally spread across conditions. There’s no reason to expect that all of the more intelligent people (e.g.) should end up in one condition. Therefore, if the only difference between your groups is the independent variable, and you find differences in the dependent variable, those differences in the DV must have been because of the IV. [Related to this, you could bring up the concept of confounds. For example, what if all the men had been given the concrete words and the women had been given the abstract words. If there were differences on the DV, we are unable to tell whether they are because of the word type or because of natural gender differences in memory.]
Note: Past research has shown that people’s memories are indeed better for concrete words. You shouldn’t panic if you are unable to replicate this result with your class. This is a rather watered down version of this memory comparison, and typically larger sample sizes are needed to demonstrate the effect. Remember that the purpose of this exercise was not to replicate the effect but to demonstrate to the students HOW one would answer a psychological question scientifically.
Part 2—Social Loafing
Personally, I think the previous exercise (concrete vs abstract memory) is a better one, and it’s sufficient to stand on its own. I think it’s more straightforward, clearer to the students, and easier to implement. You could easily do only Part 1. ...But the following is another exercise that serves the same purpose (using the scientific method to answer a psychological question). Suppose that we want to test if people work harder alone or in groups. How can we answer this question scientifically?
Again, we first define our independent and dependent variables. Our independent variable is work context (alone vs group). The dependent variable is amount of output produced by each person.
[Again, if you want to get into the nature of hypothesis testing, the null hypothesis is that there will be no difference in output as a function of work context; our alternative hypothesis is that people working alone will be more productive that people in groups.]
The instruction page (WORD97, WORD95, MacWORD5.1, WordPerfect for Windows 5.x, Rich Text Format) contains directions to the students which assign them to the alone or group condition. After making photocopies, cut the page and give one half of the class one list, and half the other. You should ask the students to not speak once they’ve received the handout (so that condition assignment information is not shared). Announce a letter of the alphabet, give them 30 seconds to produce words, and make sure they all immediately stop writing when time is up. Collect the lists and score each person’s list (DV = number of words generated). Again, you can compare means or do a t-test.
It is still vital that students be randomly assigned to conditions. Randomly split the class in half to determine who gets which list.
I’ve made instructions that you can start with, but you will need to adjust the "group" directions to describe whatever group assignment you’ll be using. Most instructors have 5-person subgroups already set up in their classes; these would be ideal to use—though possibly difficult to describe if the subgroups don’t already have group names or numbers. This is another reason why I think the memory exercise is a better one. If you don’t have subgroups, you can create arbitrary groups (men vs women, last names A-L vs M-Z, etc.). Keep in mind that this is for the instructions only, not the manipulation of the IV. You won’t actually be comparing the different subgroups’ scores, you’re randomly assigning half the class to think their output will be blended in with the rest of their subgroups’ scores.
Note: Past research has shown that people tend to work harder when they are alone; individual output decreases when the individuals are put into groups. Again, you shouldn’t panic if you are unable to replicate this result with your class. This is not the best operationalizations of both the IV (individual vs group) and DV (output) are less than ideal, and typically larger sample sizes are needed to show the effect. Remember that the purpose of this exercise was not to replicate the effect but to demonstrate to the students HOW one would answer a psychological question scientifically.