The purpose of this activity is to help you learn about design choices experimenters have and to think about the benefits and limitations of using each design. You will also learn about terminology for extraneous variables and other controls. It is designed to be studied during the Quantitative Methods unit (Chapter 6, 6.1b). It should take about 15-20 minutes.
- What are common experimental designs and controls?
- How and why are controls used in experimental research?
- When might it be impossible to use some of these controls?
- Textbook lesson 6.1b, pg 312-313
Read the summaries of the research aims below and working with a group decide which experimental design you would choose and why.
Here are your three choices for experimental design:
- Repeated measures: this is when all participants in the experiment experience all conditions of the experiment.
- For example, in a study on the effects of testosterone on brain activity, all participants would have injections of testosterone and their brain activity measured. They all would then also have a placebo treatment on a different day. Thus, the procedures are repeated on each participant.
- Independent samples: this is when you divide your participants into different (independent) groups (this design is also called independent groups) and they receive only one condition of the experiment.
- For example, in the testosterone experiment mentioned, you could divide your sample in half and one group would receive an injection of testosterone and another group would receive an injection of a placebo. Thus, the groups are independent of one another.
- Matched pairs: this is a type of independent samples that requires matching your participants on a particular characteristic that you think might be an extraneous variable. You put one participant in one condition and their matched pair in another.
- For example, perhaps you have reason to suspect that “aggressiveness” might be a variable other than testosterone that could affect brain function in your experiment. You could conduct a number of tests to give each participant a score of their level of aggressiveness. You would then match them with another participant who had the closest score and then one would receive the testosterone treatment and the other the placebo.
Read the following summaries and decide which experimental design you would choose and explain why. More information on these studies can be found in the textbook or on this blog if you are not familiar with them.
#1 Note Taking
- The aim of this study is to see which type of note-taking is more effective for long-term memory of information, taking notes with pen and paper or on a laptop. You have a group of high school participants as your sample.
#2 Misinformation effect
- This is a replication of Loftus and Palmer’s (1974) experiment on the effect of leading questions on the speed estimates of cars in an accident. You want to find out if the verb used in a critical question affects the speed your participants estimate the car to be travelling at. You have a group of students about to sit their first driving tests as your participants.
#3 A Bobo doll experiment
- This is a replication of Bandura’s famous experiments that test the effects of observing violent behaviour on the behaviour in children. You have participants aged 3-6 years old from a nearby kindergarten. You’re going to have two conditions in your experiment: observing a violent adult and a control condition where the kids don’t observe anyone at all.
#4 Laundry schema study
- You’ve decided to replicate Bransford and Johnson’s classic (1972) experiment on the effects of prior knowledge on comprehension. You have a group of international high school students from a range of nationalities as your participants. You’re going to have two conditions in your experiment – title before and no title and you want to see which one will have a bigger effect on the comprehension of the passage.
#5 Length of short-term memory
- This study is a replication of Peterson and Peterson’s (1959) study that aimed to test the duration of short-term memory. Participants are asked to memorize trigrams (a group of three consonants, e.g. MGT, PLR, KFB, etc.). Participants are read three trigrams and in one condition they repeat them straight away with no delay. In another condition they repeat after 18 seconds delay. In the original, after 18 seconds there’s almost no memory of the trigrams and you want to see if these results can be replicated.
One method of choosing a design is to first think about possible extraneous variables that might confound your results. Once you have these identified, you can then decide which design might be best for controlling for this variable.
Another approach could be to choose one of the design types first and then run through the experiment in your mind and think about potential problems with the use of this method.
Fast Finishers Extension #1
If you have chosen your design types appropriately, there’s probably a particular term for the extraneous variable you have controlled for. Using the textbook (pg. 312-313), see which of these variables you have controlled for with your choices:
- Participant expectancy effect
- Order effects
- Participant variability
Fast Finishers Extension #2
Using the same section of the textbook, how could you use these additional controls in one or more of the experiments above?
- Single-blind/double-blind design
There are lots of new terms in this lesson and do not feel overwhelmed if you cannot get them all right away. But do take careful note of this section of the textbook because it will come in very handy when you are designing your IA and writing up your report.
Think you know all the terminology in this lesson? Try this crossword puzzle to see how much you’ve learned.