Stats Flashcards
(27 cards)
Graphical methods of presentation(x5)
- scatter plot
- bar chart
- histogram
- mean +/- s/d
- box plot
When would the median and mode be different to the mean value?
Small samples and skewed distributions (non-normal)
When can the t-test be employed?
When the data (raw/transformed) follow a normal distribution
Response if P
Difference is significant – reject null hypothesis
Response if P>critical value
Cannot reject null hypothesis
Box and whisker plot: what are the values represented within the box?
Values between the 25th and 75th percentile
Box and whisker plot: what do the whiskers represent?
Whiskers show the range of data
What is the type I error?
False positive result; rejection of the null hypothesis when it is true
In what instances is it likely to cause a type I error?
Rejection of null hypothesis when p-value is close to 0.05
What is type II error?
Not rejecting the null hypothesis when it is false; might erroneously conclude that no effect is present when there actually is
Factors affecting the power of a stats test (x4)
- sample size
- variability of observations
- effect of interest
- significance level desired
Which test is used to avoid the problem of multiple corrections?
Global test - ANOVA (analysis of variance)
What do you do if the ANOVA test yields a significant result (ie if P
Move to other tests e.g. Post-hoc tests - bonferroni, scheffe
When is non-parametric stats used? (X2)
- when normality cannot be assumed
- categorical data
How is the Mann Whitney test used?
The individual data scores are converted into ranks for any subsequent analysis
What is the non-parametric alternative to the paired t test
Wilcoxon signed rank test
What is the global test used in non-parametric stats?
Kruskal-Wallis
What is the correlation analysis?
The degree of association between two variables
How is data presented in correlation analysis?
Scatter plot
What’s on the x and y-axis of the scatter plot?
X-axis: independent variable
Y-axis: dependent variable
What does pearson’s correlation coefficient measure?
The degree of linear association between two variables
Range of values of the pearson’s correlation coefficient
-1 to +1 (sign indicates the direction of association)
In what cases will Pearson’s ‘r’ value be misleading? (X3)
- non-linear r/s
- presence of outliers
- presence of subgroups
When do you not use pearson’s? (X3)
- if one variable is on the ordinal scale
- if x or y are not normally distributed
- if the sample size is too small
