data analysis Flashcards
what is a p value?
the calculated probability that there is no genuine difference between the groups. basically the strength of avidence against the null hypotheisis
if p is larger than 0.05
difference is NOT statistically significant, reatain the null hypothesis
is p is smaller than 0.05
difference is statistically significant, reject the null hypthesis
a smaller alpha value ie 0.01 is likely going to lead to a type two false ?
false negative
what is a false negative
accepting a null hypothesis when there actually IS a real effect
a larger alpha value ie 0.1 is likely going to lead to a type one false ?
positive
what is a false positive
rejecting a null hypothesis when there is NO real effect
what does parametric mean?
it adhears to normal distribution
what is quantitative data
numerical
what is continous data
can be measured
what is discrete date
only certain values ie shoe size
what is categorical data
category type of qualitiative data
what is a 95% ci
Measure of precision where 95% of the real answers lie between two given values
when would we use an unpaired t test
analysing parametric data from two independant samples
what is a mann u whitney test used for?
unpaired, independant non parametric data
non-parametric statistical test used to compare the medians of two independent groups (samples) for ordinal or continuous data.
what is a one tailed t test
One-tailed t-test: Used when you have a specific direction in mind and want to test for a difference in that direction only. For example, if you think Group A will perform better than Group B, you would use a one-tailed test.
what is a two tailed t test
Two-tailed t-test: Used when you don’t have a specific direction in mind and want to test for a difference in either direction. It’s more conservative and covers the possibility of differences in both directions.
how to we determine if a data set is parametric?
Subjectively evaluate the distribution of data (or data from previous studies) using a
histogram. E.g. is the shape correct? Is it symmetrical, skewed, what level of kurtosis is
there?
what is a paired t test
statistical test used to compare the means of two related (paired) groups for a continuous outcome variable.
When to use: When you have paired data (related samples) and you want to determine if there is a significant difference in the means between the paired groups.
ie before and after
what paired t test would you use for non parametric data
wilcoxon signed rank test
when would you use a wilcoxon signed rank test
for non-parametric statistical test used to compare two related (paired) groups for a continuous outcome variable when the assumptions for the paired t-test are not met.
what test would determine if data is parametric?
Shapiro-Wilk test)
what does anova stand for
analysis of variance
what is anova
standard approach used for statistical analysis of studies involving multiple
comparisons (and at least one continuous variable)
what is a one way anova
Studies with 3 or more varying conditions on a single continuous variable
(e.g. the effect of varying
treatments on blood pressure)
what does the first step of anova test for , what more do you need ?
first step only assesses whether there is a statistically significant difference
between any of the groups; it doesn’t indicate which groups are different (i.e. which pairs
of groups), particularly the specific pairwise comparisons that are the focus of the study
you need post tests
first step of anova only assesses whether there is a statistically significant difference
between any of the groups; it doesn’t indicate which groups are different (i.e. which pairs
of groups), particularly the specific pairwise comparisons that are the focus of the study, the tests are;
Tukey’s test is used when a study requires pairwise comparison of every possible
combination of groups.
- Dunnett’s test is used when each pairwise comparison involves one specific group (e.g.
a control group). - Bonferroni’s test (or “Bonferroni correction”) is used when the specific pairwise
comparisons required do not follow a particular pattern.
what does the anova test assume?
data is parametric and independant (unpaired)
what tests would you use for categorical data anyalysis?
McNemars test
Fishers exact test
Pearsons chi-squared test
