Placebo is normally distributed because p = 1.00 (Shapiro-Wilk because N < 30), which is more than .05
Drug A is normally distributed because p = .07 (Shapiro-Wilk because N < 30), which is more than .05
Drug B is normally distributed because p = .08 (Shapiro-Wilk because N < 30), which is more than .05

Homogeneity of variances is met because p = .23, which is higher than .05

One-Way ANOVA full report

a One-Way ANOVA was conducted to analyse the effectiveness of two treatments at reducing anxiety levels when compared to a placebo condition (Placebo, Drug A, Drug B). The assumption of normality was met upon inspection of histograms and Shapiro-Wilk (all p’s > .71). Homogeneity of variance was also assumed (Levene’s F(2, 57) = 1.52, p = .23). Results revealed there was statistically significant difference between the three conditions o mean anxiety scores (F(2, 57) = 39.80, p < .001)

Computer Lab 3

There was a statistically significant large effect of different drug types on the karaoke performance of the first university, F(2,57) = 8.43, p < .001, η² = .23, ω²=.19

There was a statistically significant large effect of different drug types on the karaoke performance of the second university, F(2,57) = 24.83, p < .001, η² = .47, ω² = .44

Additional tasks

Alpha represents probability of making a Type I Error

Beta represents probability of making Type II Error

The calculation of statistical power is 1 - β

When we observe increases in effect size, statistical power increases

When we use larger sample size, statistical power increases

When the error variance in our data increases, statistical power decreases

Alpha = 0.05, Beta = 0.3, Power = 1 - 0.3 = .70

Alpha = 0.01, Beta = 0.7, Power = 1 - 0.7 = .30

Alpha = 0.001, Beta = 0.45, Power = 1 - 0.45 = .55

From lowest to highest: BCA

Study Flaws

The study outlined in the worksheet has a number of flaws, with the biggest one being relying on self-report for assessment of song quality. Not only is self-report not very reliable on its own, the unreliability of it is exacerbated by the fact that the study’s independent variable - taking or not taking drugs - has a direct effect on the mood, and therefore, likely on the way participants perceive their own singing. Consequently, the effect may be artificially enlarged by the changed perception of own singing, rather than an actual feasible change in singing quality.

Since the study’s object is specifically on the quality of karaoke performance, the measure of it should be more objective. One of the options would be to use experts rating to asses the quality of singing, or something more technological, like measuring the proportion of notes correctly sung.

Furthermore, the study requires participants to sing a very lengthy song, which could result in fatigue, which could decrease the performance with time.

Note