Week 6 Flashcards
What is the simplest way to describe categorical variables?
Labels
- no order
- no value
What are levels in categorical variables
They have categories with different levels (e.g eye colour - blue)
What kind of variables do many psychological and health-related research questions often use?
Categorical Variables
drugs, experimental groups, sex, medicines are what kind of variables?
Categorical
What makes ordinal variables different to categorical variables?
They have order and levels.
What type of variable are those in a Likert Scale?
Ordinal
When using categorical data, what kind of questions are usually asked?
Questions about proportions. (e.g what genre was the most popular in the hottest 100?)
Give an example of factorial design using two or more variables (categorical data, stairs)
Does using the lift or stair depend at all on what direction they’re going (up or down)?
Define the degrees of freedom
In statistics, the number of degrees of freedom is the number of values in the calculation of a final statistic that are free to vary.
Give an example of a degrees of freedom
Three numbers sum to 30. How many degrees of freedom are there? First step can be random, second step can be a random amount of the remaining length but the last step HAS to cover the remaining length. Therefore it is 2. 2 degrees of random choice.
What are inferential statistics?
Statistics that answer questions about whether statistics from a sample generalise to a population.
Inferential statistics: what are p-values?
Used to make probabilistic decisions
Inferential statistics: how are p-values used to make probabilistic decisions?
Using null hypothesis statistical testing. If the p-value is small enough, reject the null hypothesis (p < .05). If p-value is bigger than criterion, fail to reject the null hypothesis.
Why do we say “fail to reject the null hypothesis” instead of keep the null hypothesis?
For clarification
What is effect size?
A (usually standardised) measure of the strength of relationship or magnitude of difference among a set of statistics.
What does effect size allow?
An assessment of practical significance (p-values only assess statistical significance).
What is a univariate test?
A chi-square goodness of fit test is one. Tests one variable
What does a chi square goodness of fit test test?
whether proportions within levels of one categorical variables are inconsistent with hypothesised portions.
What is the default assumption for null hypothesis in chi-square goodness of fit testing?
That the population is uniformly distributed across categories. Proportions are equal for all levels of the variable.
If zodiac sign is NOT associated with business uses, the number of CEOs in each birth sign should be what?
equal (uniformly distributed)
What are three chi-square data assumptions
- Data must be counts (categorical or ordinal variable)
- Counts in each category (cell) must be independent
- Sample size must be large enough (expected frequencies, or counts, in all cells should be at least 5
Discuss the things we would need to deem something having no effect in terms of the zodiac example, chi-square goodness test with 256 participants
:
N= 256 participants across all categories (example)
- uniform distribution of counts: 256/12= 21.3 per birth sign
- uniform distribution of proportions: (%) 21.3/256 = 8.35 % per birth sign
What formula is the chi-square similar to?
The formula for variance
What does chi-square measure?
How well the data match (fit) the expected distribution.
What does a small chi-square indicate?
A good fit. 0= perfect fit
What is the problem with chi-squares?
- magnitude of chi-square statistic is dependent on
- sample size
- degree of freedom
How do you calculate the degree of freedom in a goodness of fit test?
df= cells -1
In regards to Cramer’s V, what does a score of >5 equal?
A large effect
In regards to Cramer’s V, what does a score of .3-.5 equal
medium effect
In regards to Cramer’s V, what does a score of .1-.3 equal
small
In regards to Cramer’s V, what does a score of 0-.1 equal
trivial
What range is the Cramer’s V?
between 0 and 1 (always positive because to find it, you have to square the number)
What do you click in Jamovi to begin a goodness-of-fit test?
Frequencies-one sample proportion tests- N outcomes
Can you do a chi-square test if the expected count is lower than 5?
NO. Collect more data, and if you can’t then check data cautiously
What should we do if in a chi-square goodness of fit test, the p value is greater then .5?
Fail to reject the null, effect size is not statistically significant
When is the chi-square test of independence used?
To test whether two categorical variables are independent. (e.g is choice of major independent of biological sex?)
What is an example of a chi-square test of independence using the titanic as an example?
Was survival on the titanic independent of ‘class’ of passenger? IF proportions are independent, all levels of the other variable will equal the total proportion.
How do we calculate expected frequencies?
Expected Frequency = (row total X column total)/N total
For chi-square test of independence, how do we calculate the degrees of freedom?
df=(#rows-1) X (#columns - 1)
df= (R-1)X(C-1)
In test of independence, what makes the test robust in relation to cell frequency?
Robust if fewer than 20% of cells have small expected frequencies. If expected frequencies are small (<5), observed frequencies cannot be normally distributed about expected frequencies.
COLLECT MORE DATA.
What are the steps to do a test-of independence with Jamovi?
Frequencies, independent samples, x^2 test of association,