A study was conducted to investigate the effectiveness of a new anti-anxiety drug; the effects of the drug were tested on participants that either had low or high baseline levels of anxiety. The researchers hypothesized that the drug would be more effective at reducing anxiety in individuals with higher levels of baseline anxiety. The researcher tested a total of 48 participants, with 12 assigned to each of four experimental groups resulting from a factorial combination of the two independent variables. The effects of the drug were assessed using a well-validated self-report questionnaire, which measured anxiety on a 20-point scale with higher scores indicating greater levels of anxiety. 45 minutes before anxiety measures were taken, the participants in the study were administered either the anti-anxiety drug in pill form, or an identical placebo pill. Pills had different colors so that the experimenter could easily identify what they contained. Participants were informed in the consenting protocol that the study involved a new anti-anxiety drug, but were naïve regarding the aims of the study or the color coding of the pills. On debriefing, participants were approximately 50% correct when asked which experimental group they belonged to. To analyze the results the researcher used a two-factor ANOVA each with two levels: baseline anxiety (low vs. high); drug condition (placebo vs. anti-anxiety drug). The following table displays the mean anxiety scores post-drug for each condition.
|
Drug Condition
|
|
Baseline anxiety
|
Placebo
|
Anti-anxiety drug
|
|
Low
|
4
|
18
|
|
High
|
14
|
17
|
a. The main effect of drug was significant. Referring to the table of means, how would you describe verbally the effects of the drug?
b. The 'interaction effect' was also significant. Using the means in the table above write a statement describing how the anti-anxiety drug appears to affect anxiety. Are the results consistent with the hypothesis proposed by the researchers?
c. What aspects of the design help to control differences in demand characteristics between conditions, and what possible confound is still present? What method would be appropriate to rule out this above confound and how might it be implemented here?