Advanced Experimental Design

• Memory test 1 SLEEP: bed, rest, awake, tired, dream, wake, snooze, blanket, doze, slumber, snore, nap, peace, yawn, drowsy
• Memory test 2 WINDOW: door, glass, pane, shade, ledge, sill, house, open, curtain, frame, view, breeze, sash, screen, shutter ... banana
• Independent variables?
• Dependent variables?
• Design?

• More than one independent variable
• Each level of one independent variable is paired with each level of every other independent variable

• 2 X 2: two variables with two levels each
• 2 X 2 X 3: three variables, 12 conditions

• Main effects
• The separate effect of each independent variable averaging over the other variables.
• E.g., Younger people do better than older people.
• Interactions
• The effect of an independent variable differs depending on the level of another independent variable.
• E.g., Younger better than older for inconsistent and target but not for consistent.

• Main effects
• Coffee
• Time of day
• Interactions
• Crossover interaction is the strongest.
• Floor/ceiling effects and interactions.

• Main effect of factor A
• H_A0: Mean_A1 = Mean_A2
• H_A1: Mean_A1 < > Mean_A2
• Main effect of factor B
• H_B0: Mean_B1 = Mean_B2
• H_B1: Mean_B1 < > Mean_B2
• Interactions
• H_I0: Mean_A1B1 - Mean_A2B1 = Mean_A1B2 - Mean_A2B2
• H_I1: Mean_A1B1 - Mean_A2B1 < > Mean_A1B2 - Mean_A2B2

• Threats to internal validity - Does my experiment test what it is supposed to?
• Selection: Differences across groups may be due to pre-existing differences in the participants.
• Maturation: People's responses can change from the beginning of the study to the end in ways that are unrelated to the independent variable.
• Mortality: Some participants may drop out of the study, leading to group differences because of changes in the sample rather than because of the independent variable.
• History: An event that occurs outside the context of the research may affect one group, causing a difference that appears to be related to the independent variable but actually is not.
• Instrumentation: Change the way the dependent variable is measured may cause a difference in the dependent measure.
• Testing: When people undergo repeated testing, they may change because of the process of testing rather than because of an independent variable.
• Statistical Regression: Scores on an initial testing may be extreme and regress to the mean, leading to the appearance that scores changed because of the independent variable.
• External validity - Can I generalize to the real world?
• Is my sample representative?
• Is there a point to my research?

1. Identify the design. How many independent variables are there and how many levels does each one have?
2. Identify the total number of conditions.
3. Identify the independent variables.
4. Is this a repeated measures design? If so, identify the repeated variables.
5. Identify the dependent variables.
6. Identify problems (threats to internal validity and external validity).

• The researchers wanted to know if cell phone use might impair a person's ability to operate an automobile. They asked participants to use a joystick to move a cursor across a course on a video screen and to press a button whenever they say a red light appear on the screen. The participants were randomly assigned to conditions in which they 1) listened to the radio (a station of their choice), 2) communicated with a confederate by means of a hand-held cell phone, or 3) communicated with a confederate by means of a hands-free cell phone. The participants engaged in the tracking task for 7.5 minutes, then engaged in a dual task of the tracking while 1), 2), or 3). The researchers discovered that most participants detected all of the red lights; the percentage of the time that they missed the signal was small. The participants using cell phones were significantly slower in responding to the signal when engaged in conversation than when they were only doing the tracking task. This was not the case for the participants listening to the radio, who performed with equal speed in spotting the red light in the single and dual tasks. Based on this study and a related experiment they conducted, the investigators concluded that operating a cell phone while driving is likely to lead to poorer driving. It does not matter if the cell phone is hand-held or hand-free. Listening to the radio did not affect performance.

  Strayer, D. L., & Johnston, W. A.(2001). Driven to distraction: Dual-task studies of simulated driving and conversing on a cellular phone. Psychological Science, 12, 462-466.

• Two psychologists investigated the relation between playing action video games and the ability to monitor the visual environment. They recruited a group of male undergraduates who played action video games at least one hour a day, four days a week over the previous six months. A control group of non-video game players also participated in the study. The researchers discovered that video game players were better at monitoring visual elements unrelated to the video game while they were playing the game. The researchers concluded that playing video games increases oneีs capacity to pay attention to details in a visual environment.

  Green, C. S., & Bavelier, D. (2003). Action video game modifies visual selective attention. Nature, 423, 534-537.

1. Identify the design and total number of conditions.
2. Identify which variables are manipulated variables, measured variables (i.e., participant variables), and repeated measures variables.
3. For each IV separately, identify the means of each level of that IV.
4. The interaction is significant. Interpret the data. What do the results tell you?

The "Mozart Effect" suggests that spatial-temporal abilities improve after listening to music composed by Mozart. Participants agreed to attend to two separate testing sessions in the lab. In the first session, participants listened either to music composed by Mozart or to a story that had been recorded on tape. Then they took a test that included 17 questions. The participants then left the lab, returning after two weeks. If they had listened to Mozart during the first session, they listened to the story this time; if they had listened to the story during the first session, they listened to Mozart this time. The researchers recorded whether the participants preferred the story or Mozart. The participants then completed a second test of 17 items.

Table 1. Mean number of problems correctly answered on the 34-item tests.

Preference	Sample size	Listen to Morzat	Listen to a Story
Mozart	13	14.62	13.23
Story	15	11.60	12.67

1. What are the IVs and DVs in your project?
2. What are the operational definitions of your variables?
3. What is the design of your experiment?
4. What are the levels of your IVs?
5. What are the potential problems?
6. How are you going to analyze the results?