This is my ultimate statistical guide. It is just a snapshot of a stat summary... It is about how to select statistical tests. Test1: test if the distribution are normal

shapiro.test(score) If p-value > 0.05 the distribution is not normal

(If not normal, try with the log(score) or with the sqrt(score) : shapiro.test(log(mm)), data are transformed and the normality test is performed another time. If still not normal, try a non-parametric test : kruskall or wilcox)

Test2: variance homosedasticity (to see if variances are equals) If 2 groups: var.test If more thant 2 groups: bartlett

Here: 2 groups (with at and without): var.test(score ~ awareness) The first (score) is the quantitative indicator and the second (awareness) is the factor. Here p-value = 0.492 > 0.10 then variances are equals. That's what we want to perform the variance analysis

(If the homosedasticity test says that the variances are not equals or if the distributions are not normal: wilcox test (2 modalities) or kruskall test (more than 2)

Now we want to test the impacts of the awareness tool on score Since we have ONE factor with TWO modalities, we can perform a STUDENT TEST (if more than 1 factor: ANOVA)

t.test(score ~ awareness, var.equal=T) p.value= 0.04276 . H0 (equals means) is rejected then there is an effect of awareness on score

anova(lm(score ~ awareness)) (if F< 1 we reject in any case of pvalue)