DAPR 1 content

Question 1

what is the null hypothesis?

Answer 1

H0 = the status quo, what is currently accepted or ‘no change’

Question 2

what is the alternate hypothesis?

Answer 2

H1 = encodes our research question suggests change (can be directional - one tailed/two tailed)

Question 3

what is µ (mu)?

Answer 3

represents the mean for an entire population

Question 4

what is a statistic?

Answer 4

any number which summarises data in a sample e.g. mean or sd

Question 5

what is a parameter?

Answer 5

any number that describes a whole population - they are usually unknown values that we have to estimate

Question 6

What is a sampling distribution?

Answer 6

collect many samples and plot the means of the resamples on a histogram.

Question 7

what is a t-distribution?

Answer 7

a mathematical curved line that fits the distribution of the sampling distribution.
it is dependent on the degrees of freedom.

Question 8

what is standard error?

Answer 8

SE is the standard deviation of the sample mean

Question 9

what is the equation for SE?

Answer 9

SE = standard deviation of the original sample / square root of n

Question 10

what is a t-statistic

Answer 10

a value describing our data based on the t-distribution
the larger the t-statistic, the more likely we are to reject the null

Question 11

what is a p-value?

Answer 11

a measure of the strength of the evidence that the data brings against the null hypothesis.

Defined as the probability of obtaining sample results as or more extreme than the ones observed if the null was true.

If our p-value is smaller than our chosen significance level, this suggests the results are not just due to ‘noise’

Question 12

What is the difference between t-statistics and p-values?

Answer 12

t = determines if variation in data is due to random sampling or not

p = measure the significance of the evidence against the null hypothesis

Question 13

what is a confidence interval?

Answer 13

provides a plausible range of values (an interval) you are confident will contain the unknown parameter - if your null value is within the interval we FAIL to reject it

the interval chosen is related to the significance level e.g. 5% alpha = 95% confidence interval

Question 14

what is effect size (Cohen’s d)?

Answer 14

determines whether the difference between observed and hypothesised values are important (this is different from significance)