Slides

PsychExchange.co.uk Shared Resource

Description
1. Mann Whitney U Test <br />A test of difference<br /> 2. A Tip<br />If there is an R in the stats test name then the <br />critical value must…
Categories
Published
of 19
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  • 1. Mann Whitney U Test <br />A test of difference<br />
  • 2. A Tip<br />If there is an R in the stats test name then the <br />critical value must be gReateR than the <br />observed value in order to reject your null hyp<br />No R = observed value must be less than critical <br />value to reject your null hyp<br />
  • 3. Decision 1: What are you looking for?<br />Differences<br />Correlation<br />Spearman’s rho<br />Decision 2: What level of measurement do the data represent?<br />Nominal<br />Ordinal (or interval)<br />Chi squared test<br />Decision 3: What exp design was used?<br />I.G<br />Related<br />Mann Whitney U Test<br />Wilcoxon<br />
  • 4. Inferential Tests<br />Which test to use depends upon a number of<br />factors:<br />The type of data: Ordinal <br />Type of research design (RM vs. IG)<br />One tailed or two tailed test<br />
  • 5. Inferential Tests<br />Which test to use depends upon a number of<br />factors:<br />The type of data: Ordinal (or ranked from interval)<br />Type of research design (RM vs. IG): IG (Wc uses RM)<br />One tailed or two tailed test<br />
  • 6. Mann Whitney U Test<br />Uses when we are looking for a difference eg, if <br />a hyp states:<br />“There will be a difference between the <br />performance of Cats and Dogs on an IQ test”<br />Or <br />“Dogs will score higher than cats on an IQ test”:<br />Null hyp: “there will be no difference between the <br />performance of cats and dogs on an IQ test”<br />
  • 7. A worked example<br />YOU DON’T NEED TO KNOW THE MATH<br />But sometimes its nice to know!<br />
  • 8. Alternative Hyp/Exp hyp:<br />More words are recalled in the exp condition <br />than in the control condition<br />1 or 2 tailed hyp?<br />Null hyp?<br />
  • 9. 1 Place data into table. <br />2 Rank data lowest value to (rank 1) highest value (rank N*) Split values for those that share the same<br />3. Calculate the sum of the ranks for the smaller of the two samples (if they are both the same size calculate the sum of the ranks for either) this value is T<br />
  • 10. A memory experiment<br />
  • 11. Formulae - Step one<br />N₁N₂ + (N₁ + 1) - T<br /> 2<br />U = observed value of MW<br />N₁ = number of values in the smaller sample (or the sample for which sum of rank calculated)<br />N₂ = number of values in larger sample (or sample for which sum of ranks not calculated)<br />
  • 12. N₁N₂ + (N₁ + 1) - T<br /> 2<br />9x10 + 9x (9+1) – 47<br /> 2<br />Blah blahblah<br />U= 88 this is the observed value of U<br />
  • 13. Formale Step 2<br />U’ = N₁N₂ - U<br />Blah blahblah<br />U = 2<br />Select the smaller of the 2 values out of U and U’<br />The smallest value become U<br />
  • 14. Critical Value of U is the maximum value of U <br />that is sig at a certain level. <br />To work it out you need N₁ N₂ ( 9 and 10)<br />Critical value must be equal to or less than the <br />observed value of U to reject that null <br />hypothesis<br />
  • 15. Calculate<br />Boys will rate chocolate as tastier after playing <br />computer games<br />N₁ 8 N₂ 9<br />Observed value of U = 19<br />*What would the design of the experiment be?<br />*Is it a one or two tailed test? State how you know?<br />*What would the null hyp be?<br />*Is the test significant and at what level? State what <br />the critical value should be and what it was and explain if you <br />would then expect or reject your null hypothesis<br />*What is the chance of a type II error?<br />
  • 16. Calculate<br />There is a difference between how attractive women find <br />men dependent upon whether or not the woman had <br />cheese for lunch<br />N₁ 12 N₂ 9<br />Observed value of U = 21<br />*What would the design of the experiment be?<br />*Is it a one or two tailed test? State how you know?<br />*What would the null hyp be?<br />*Is the test significant and at what level? State what <br />the critical value should be and what it was and explain if you <br />would then expect or reject your null hypothesis<br />*What is the chance of a type II error?<br />
  • Neww

    Jul 23, 2017
    We Need Your Support
    Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

    Thanks to everyone for your continued support.

    No, Thanks