Week 4 - Statistical models (correlations and t-tests) Flashcards
Error in statistical models…
a. Purely reflects measurement error
b. Is only apparent on rare occasions
c. Is typically ignored
d. Is important to measure, report and interpret
d. Is important to measure, report and interpret
What three factors affect statistical power?
a. Sample size, effect size, type 1 error rate
b. Sample size, statistical significance, type 1 error rate
c. Sample size, statistical significance, independence
d. Sample size, effect size, independence
b. Sample size, statistical significance, type 1 error rate
Data = x + y, where x and y are…
a. Model and error
b. Distribution and model
c. Variance and error
d. Variance and distribution
a. Model and error
What is not a step in Null Hypothesis Testing?
a. Formulate an interpretation of the data before analysis
b. Assess the “statistical significance” (p value) of the result
c. Collect some data relevant to the hypothesis
d. Formulate a hypothesis that embodies our prediction
a. Formulate an interpretation of the data before analysis
The probability (p) value is…
a. The probably that if you ran the study again, you would obtain the same result that % of the time
b. The probability of the null hypothesis being true
c. The probability that you are making the wrong decision
d. Under the assumption that the null hypothesis is true, the probability of getting a sample as or more extreme as our own
d. Under the assumption that the null hypothesis is true, the probability of getting a sample as or more extreme as our own
What is an appropriate way to report a statistical model?
a. The effect approached significance
b. There was a trend for significance
c. The effect was nearly significant
d. The effect was statistically significant
d. The effect was statistically significant
The alpha value is…
a. Set before analyses are run
b. Akin to the p value
c. Always 0.05
d. An alternative measure of effect size
a. Set before analyses are run
Type 1 errors reflect x, and type 2 errors reflect y. What are x and y, respectively?
a. They both reflect false positives
b. They both reflect false negatives
c. False positives and false negatives
d. False negatives and false positives
c. False positives and false negatives
Effect size is a measure of…
a. The significance of the effect
b. The variability of the effect
c. The error of the effect
d. The strength of the effect
d. The strength of the effect
Statistical significance is not…
a. Determined by comparing the p-value with the alpha value
b. Based on a relatively arbitrary cut-off point
c. A dichotomous assessment
d. A measure of effect size
d. A measure of effect size
Which test is parametric?
a. Pearson correlation
b. Spearman correlation
c. Wilcoxon test
d. Kruskall-wallis test
a. Pearson correlation
In relation to the hypothesis, ‘In a mirror drawing task participants will trace a star figure faster (measured in seconds) on the trials in which they used their dominant hand than on the trials in which they used their non-dominant hand’
What is the inferential statistical test of choice?
a. Paired samples t-test
b. Independent samples t-test
c. Pearson Correlation
d. ANOVA
a. Paired samples t-test
A statistical model uses ____ to summarise a dataset relative to ___:?
a. Statistics, a single variable
b. Maths, a single variable
c. Statistics, multiple variables
d. Maths, multiple variables
d. Maths, multiple variables
in relation to the hypothesis, ‘Intellectually disabled adolescents given different types of skills training programs (Group 1=Social Skills, Group 2=Daily Living Skills, Group 3=Both Social Skills and Daily Living Skills) will differ in their level of Assertiveness (measured on the Goldberg Assertiveness Scale).’
What is the inferential statistical test of choice?
a. Oneway ANOVA
b. Independent samples t-test
c. Pearson Correlation
d. Paired samples t-test
b. Independent samples t-test
You should construct a scattergram before conducting a Pearson correlation because
a. it tells you whether the result is significant or not
b. there is no point in computing the correlation unless you can see a clear line formed on the graph
c. there are different procedures depending on whether there appears to be a positive or a negative correlation
d. you should not conduct a Pearson correlation if you see a clear curvilinear relationship
d. you should not conduct a Pearson correlation if you see a clear curvilinear relationship