Intro/general

Question 1

Diff between within and between subjects design

Answer 1

Repeated measures- within subjects . (Each has multiple data points, no indv diffs, cost and time effective, but has order effects) Independent- between subjects (one data point each, no order effects, takes longer and has individual diffs).

Question 2

Matched pairs design

Answer 2

Diff ps in each condition, ps matched to account for individual diffs. Best of both but hard to match accurately . One p per condition

Question 3

Confounds vs extraneous variables

Answer 3

Extraneous: not controlled in experiment, could have effect on dv (weather, noise). Confounding are extraneous variables that VARY SYSTEMATICALLY WITH THE IV TO INFLUENCE THE DV are likely to influence the results (e.g. one condition longer and more boring)

Question 4

Levels of data

Answer 4

Categorical data: nominal and ordinal, also called discrete . Ordinal are categories with a hierarchy like illness stage, but distance between scale points aren’t always equal. Continuous data are interval (no meaningful 0, intervals area equal like temp) or ratio. (Has meaningful 0 and intervals equal like time)

Question 5

Descriptive stats

Answer 5

Numbers that summarise data: mean, median, mode, SD, standard error, range . Can present in text, tables or graphs . For standard error- SD/ square root of N

Question 6

What they mean and how to calculate the SD and variance

Answer 6

Both show how spread the data is from the mean. Start W variance: find mean per condition. For each P, subtract mean from there score (d). Do d squared. Sum of all d squared for each condition then divide sum by n-1 . Then square root of variance for SD

Question 7

What is a Z score and how to calculate

Answer 7

A standardised score that represents a datapoints relationship to mean of group. Z= (score-mean) divided by standard deviation . 1.96 is a z score

Question 8

How to present the levels of data

Answer 8

Nominal can be presented as frequencies, continuous should be as means or SD in text, graph or table (mean= seconds, SD= )

Question 9

What does normal distribution mean

Answer 9

95% data under normal curve fall within the interval 1.96- a z score

Question 10

What are confidence intervals and how to calculate

Answer 10

Used as a measure of the spread of data, has an upper and lower number and 95% confident pop mean lies within this range . Confidence intervals= mean plus or minus (1.96*standard error)

Question 11

How to work out the sampling error and how to lower it and when it is higher

Answer 11

Sample mean-population mean=sampling error (mean of sample may be diff to mean of population). To lower, take an average across multiple samples. In small samples, likely to be higher as most score above or below the mean

Question 12

What is the sampling distribution

Answer 12

The sampling distribution of the mean is a normal distribution

Question 13

Diff between type 1 and type 2 errors

Answer 13

Type 1- I’m the best when you’re wrong- false positive. Type 2- I’m worthless but you’re wrong- false negative

Question 14

What is the alpha level

Answer 14

The threshold for significance: p= 0.05- means 5% prob results are due to chance. P<0.05 is significant but equal to or more is NOT sig

Question 15

How to report p values

Answer 15

To 3 decimal places, no leading 0, lower case italics p and report the exact value not more or less than 5, exception is <.001 never 0

Question 16

General format for reporting results

Answer 16

Data handling: say what numbers-raw tp change. Descriptive stats: central tendency and variance (mean and sd) or median and interquartile range . Which test and why -met assumptions . Test stat and p value