Unit 6.1 - statistical analysis

What do error bars represent? (6.1.1)

• error bars are a graphical representation of the variability

• o   Represent a description on how confident you are that the mean represents the true value

• SD error bars: reflect the spread of data around the mean

• SE error bars: indicate the precision of an estimate of the mean

For a normal distribution what is the % of values that fallen within 1,2 and 3 standard deviations (6.1.3)

• 68% of values fall within plus/minus 1 standard deviation of the mean

• 95% of values fall within plus/minus 2 standard deviation of the mean

• 99.7% of values fall within plus/minus 3 standard deviation of the mean

68 = 95 = 99.7

Define standard deviation (6.1.2)

It is a statistical analysis tool used to summarise the spread of values around the mean

(square root the variance)

Define variance (6.1.2) and give its formula

Averaged squared deviation from the mean of the set of the data

sum of the (value - mean)2 / n-1

What does a small or large standard deviation indicate (6.1.4)

Small standard deviation - indicate that the data is clustered closely around the mean value

Large standard deviation - indicate a wider spread/large range of values around the mean

**indicate possible inaccuracies in variables/experiment

Define coefficient of variation (6.1.5)

the standard deviation expressed as a percentage (%)

SDV/MEAN X100

state the 4 types of t-tests (6.1.6)

paired, unpaired, two tailed, one tailed

define a t-test (6.1.6)

A statistical test that is used to compare the means of the two groups and measure the amount of overlap between two groups

**need SS above 5 & data should be normal distribution

compare an unpaired and paired t-test (6.1.6)

paired = used when we are interested in sig difference between 2 variables for the same subject (were in both conditions)

unpaired = compare the mean of two independent/unrelated groups & to see if there is a sig difference

compare a two and one tailed t-test (6.1.6)

one tailed = test if there is direction relationship between variables (difference in a specific direction)

two tailed = non-directional relationship between variables (detect any significance difference between the means, regardless of the direction)

Explain that the existence of a correlation to causation (6.1.7)

• correlation does not imply caution/cause-effect relationship (that there is a casual relationship between two variables), cannot say it is the only factor

• correlation range from -1 to 1

• just because two variables are related, does not validate a cause-and-effect relationship between variables