Week Seven Flashcards
(18 cards)
What does the t-value measure?
The difference between the sample mean and a constant, or between two means, in relation to the precision with which they were estimated.
What does standard error measure?
The amount of chance of fluctuation or imprecision we can expect in sample estimates.
What value does the t-value to take to be seen as significant?
Greater than +2 and smaller than -2
Why do we use a one sample t-test?
When you want to compare a sample mean with a known mean.
When do you use an independent samples t-test?
When you want to compare two independent samples.
What do type 1 errors depend on?
Confidence levels. The p-value indicates the chance of committing a type 1 error.
What do type 2 errors depend on?
The level of significance and the properties of the data.
What is a one-tailed test?
Sometimes known as a directional test, where we hypotheses where the t-value is going to fall.
What is a two-tailed test?
If we are hypothesising that there is a difference between variables but not suggesting what that relationship is. It is non-directional.
What is ANOVA?
It is a way of comparing the mean of continuous variables, between three or more groups.
What is the test statistic recorded as for ANOVA?
The f-statistic.
What is the equation for the f-statistic?
F = between group variability ➗ within group variability.
How can we test whether the observed rates differ from chance?
We apply the chi-squared test.
How do we calculate degrees of freedom?
DF = (the amount of variables in the first category ➖ 1) ✖️ (the amount of variables of the second category ➖ 1)
What does chi-square measure?
Whether there is more variation across categories, in a two-way table, than we would expect to happen by chance.
What does x-squared equate?
The sum of the contribution from each of the individual cells. Every cel contributes something to the overall x-squared.
What does the significance of variation depend on in the chi-squared test?
The size of x-squared and the degrees of freedom.
What do independent sample t-tests, On my way!-way ANOVA and two-way tables all have in common?
We have one outcome of interest and it is explained in terms of one other variable.
