PAS 5

Question 1

What do error bars represent?

Answer 1

the uncertainty or variation in your data, they indicate how spread out the data is around the mean.

• They often represent the standard deviation, the standard error, the range or the confidence interval.

Question 2

give examples of Descriptive statistics and Inferential statistics

Answer 2

Descriptive statistics: range, standard deviation

Inferential statistics: standard error, confidence intervals

Question 3

What does the overlapping of error bars indicate?

Answer 3

if the error bars do overlap by quite a bit then….
it’s a clue that the difference is not statistically significant.

When standard deviation error bars do not overlap, it’s a clue that the difference may be significant

You must actually perform a statistical test to draw a conclusion.

Question 4

caution with SEM error bars

Answer 4

with low numbers of samples, they underestimate the uncertainty

error bars shows only 1 SEM (68%). by eye you should be able to get the 95% value

Question 5

What does the t-test assume and how should t-test be used if the data is skewed?

Answer 5

The t-test assumes your data:

* are independent
* are (approximately) normally distributed.
* have a similar amount of variance within each group being compared

If you know that your data is strongly skewed
For large numbers of observations, you are safe to use the t-test
For small numbers of observations, the t-test may not be accurate

Question 6

when do we use a paired t-test

Answer 6

when data can be linked in some way – for example before and after data, matched or paired data sets.

do not confuse with independent t-test

Question 7

what is a t-test used for?

when are they usually used?

Answer 7

Statistical significance i.e. is the difference down to chance or is it robust

T-tests are normally used for small groups of data up to 30 approx.

Question 8

compare inferential and descriptive data?

Answer 8

Descriptive statistics tell us what we have

Inferential statistics tell us what we could expect if we did it again

Question 9

how do degrees of freedom vary for a t-test with different group numbers?

Answer 9

• So just as the degrees of freedom were n-1 for 1 group tests, the degrees of freedom for a 2 group t-test are 2 x n -2.

n = number of samples

Question 10

why do we have figure legends?

Answer 10

figure legend of the graph states what error bars represent