Statistics Hypothesis Testing Hypothesis is a testable statement

Statistics

Hypothesis Testing Hypothesis is a ‘testable statement’ Types = alternate, research, experimental (H 1), null (H 0) They are 1 or 2 tailed (directional or non directional (but not the Null) They include an IV and a DV They are operationalised (precisely defined in terms of how the IV and DV will be manipulated or measured) Aim of research is to accept/reject H 1 or H 0 Complete the exercise

Inferential Statistics 1. 2. 3. 4. 5. Analyse using descriptive statistics (tells you whether a difference exists or not) Analyse using inferential statistics (tells you whether the differences are significant) If the sample has yielded significant results, we can infer the same is true of the population The statistical tests stringently process the data to tell you whether or not chance has caused the outcome Therefore part of the whole process involves having a null hypothesis (HO) and levels of significance or probability e. g. 5%

‘Proof’ In science it is only possible to prove something is not the case e. g. ”all swans are white” Can’t be proved But can be disproved – HOW?

Significance levels Likelihood or probability Expressed as a % and as a decimal e. g. heads or tails 50% or 0. 5 Picking the ace of hearts from 4 aces 25% or 0. 25 Likelihood of having schizophrenia 1% or 0. 01

Ruling out Chance • Standard level of significance in Psychology = 5% (0. 05) Accept H 1 – then p < 0. 05 • Accept H 0 – then p > 0. 05 • BUT a significant result might still be wrong 5 times in 100 – in other words it happened due to chance – we live with this risk •

Type 1 and type 2 errors H 0 is actually true Accept H 0 OK H 0 is actually Type 2 false You conclude there isn’t an effect when there is Reject H 0 Type 1 You conclude there is an effect when there isn’t OK

Type 1 and type 2 Type 1 is more likely when we have a high significance level e. g. 10% Type 2 is more likely when we have a low significance level e. g. 1% At 5% both are equally likely

What you need to be able to do Identify an appropriate statistical test Explain your choice State a conclusion based on a stats test Write a null hypothesis Explain why a particular stats test was used

Descriptive Statistics Descriptive statistics give us a way to summarise and describe our data but do not allow us to make a conclusion related to our hypothesis. For example, measure of central tendency such as ______, ______ or ______. It also includes graphs and charts, and measures of dispersion such as _________ or _______.

Tasks Measure of central tendency ◦ Match the definitions to the terms, and the strengths and weaknesses. Measure of dispersion ◦ What is a range? ◦ Read about standard deviation What are bar charts and scatter graphs? Define and draw an example. Complete the memory experiment tasks

What is the point? Why do we bother to use inferential statistical tests? ◦ Inferential statistics allow us to draw conclusions from findings. ◦ They allow us to see whether our results are the result of something happening, or are just down to chance. Cross out the words on the sheet and fill in the table

Pg 20 Read the example about the chip bins and female drivers. Read “Using Statistical Tests” on pg 2223 Answer the questions on the sheet

Levels of significance This refers to the minimum probability we will accept that our results are due to chance. If it is too lenient, then our results may appear to be significant when in fact they are not. If it is too stringent, then our results may appear to be insignificant when they actually they are. In psychology, we generally aim for a significance level of _______%. This means that we can be _____% certain that our results are not due to chance. This is written as P≤_______

Example Read the example about biscuits On the sheet, cross out the right words and fill in the table

Levels of measurement Nominal Ordinal Interval Ratio NOIR Read the descriptions on the sheet give your own examples

Choosing the right test Design Repeated measures Nominal data Sign test Matched Sign test pairs Independe Chi square nt measures Correlation Chi square Ordinal data Wilcoxen signed ranks Wilcoxen Mann Whitney U Interval data Related t test* Unrelated t test* Spearmans Pearsons rho product * Refers to Parametric tests (see handout)