Lecture Week 8 Flashcards
continuous/discrete = ?
continuous = can take on any value within an interval
discrete = set distinct values within a dataset
bivariate = ?
focusing on two variables at once
two group t-test = ?
comparing the means, SDs and sample size between 2 variables
alpha significance level?
5% or 0.05
should you always begin with univariate descriptives?
yes
describe one variable at a time individually first
p value = ?
the measure of statistical significance
reject null hypothesis if p value is less than 0.05/5%
fail to reject null hypothesis if p value is more than 0.05/5%
what are two descriptive tools that can be used to assess the relationships between two continuous varibles?
scatterplots & correlation
scatterplots = visualises the relationship between two variables (strength & direction)
correlation = captures strength & direction
what are outliers and what causes them?
exceptionally small or large value in a dataset
- incorrectly recorded data
- correct data value incorrectly included in dataset
- correctly recorded data genuine outlier
what is the 1.5 x IQR approach?
any potential outlier is a value any value above Q3 + 1.5 x IQR
does empirical rule (68-95-99.7) show outliers?
yes
correlation coefficient?
has the same basic purpose as scatterplot
represented as a single number
pearson’s r correlation coefficient?
r
shows how tightly the points in a scatterplot are concentrated around the straight line
only useful when relationships are linear
spearman’s rho correlation coefficient?
p
provides as much information as the pearson’s one, but with less stringent assumptions
r squared measures…
the amount of variance in the dependent variable explained by the independent variable
is showing correlation the same as demonstrating causation?
no
