week 3 - ethics and stats of MRI Flashcards
(26 cards)
Why do scientists create and design cognitive tests?
to quantify cognition and create a marker for something that is cognitive theory
Which cognitive process does each of these tests grade?
Trail Making
Letter Fluency
Trail Making: processing speed and aspects of executive function
Letter Fluency: word generation, sematic abilities
What is the difference between using parametric and non-parametric tests?
parametric tests assume the data you’re using is normally distributed whereas non-parametric is not
What values do parametric and non-parametric tests produce?
parametric: mean
non-parametric: rank, median,mode
What determines the size of the bell in a Gaussian distribution?
value of standard deviation
What is variance in a normal distribution? What does it tell you?
the average difference of all individual data points from the mean. variance shows how tightly clustered the data is around the mean
What is sd? What does it show?
How does it relate to variance?
-average distance from the mean. a measure of how dispersed the data is in relation to the mean
-sd is square root of variance
Increasing sample size increases ___________
confidence in results
What are the two common statistical designs for MRI studies?
- Groupwise comparisons
- Correlations
What test would you select if you had two groups and parametric data?
independent or paired t-test
When do you use a independent t test?
used to compare the means of two independent groups to see if there’s a significant difference between them.
How you do determine DOF for a t test?
sample size N -1
What happens to the significance threshold as N sample size increases for a t test?
threshold decreases as N increases -> easier to achieve statistical significance
what is the difference between a paired and independent t test?
independent: groups are unrelated
paired: testing same group maybe at different time points
What helps to make a t value larger -> thus grater chance of statistical significance?
Greater distance between means
Smaller SDs
Larger Ns
As the t test value increases, what happens to statistical significance?
t value increases, greater chance of reaching p value threshold (table) and statistical significance
What test would you use if you wanted to test correlation between two variables using parametric analysis?
Pearson’s
What do the Pearson’s test values mean?
-1 = perfect negative correlation
0 = no relationship at all
1 = perfect positive correlation
Does correlation mean causation?
no!
What test would you choose for non-parametric analysis, groupwise comparisons and with independent groups?
Mann-Whitney U
What test would you choose for non-parametric analysis, groupwise comparisons however using the same group?
What is the parametric equivalent?
Wilcoxon signed-rank
Paired t test
What non-parametric test would you use to show correlations?
What is the parametric equivalent
Spearman correlation
Pearson’s
Why do we use rank for non-parametric analysis?
massively reduces the disproportionate effect that extreme values have on the mean
What are the characteristics of non-parametric analysis?
skewed data
data points are ranked/use median (instead of using mean)
