s'lipkaderby'ac'uk

1

• s.lipka_at_derby.ac.uk
• Summary of group activities during last week
• Aims for todays workshop
• calculate subjectmeans
• run AOVs
• discuss poster guidelines
• Aims for next workshop

2

• Summary of activity during last week
• finish running experiment (the more participants,
  the better)
• prepare your results file for stats

3

• This week do stats in SPSS
• datafile imported Authorware results file
• --gt give your Values meaningful names
• --gt create new variable Accuracy
• 1 correct response
• 2 incorrect response
• ----gt Transform, Compute, type in
  Accuracy as Target Variable, click IF, click
  Include if Case satisfies Condition in field
  Numeric Expression type Exp_Resp Choice, click
  Continue, in field Numeric Expression type 1,
  then OK.

4

• do stats in SPSS
• --gt create new variable Accuracy contd
• ----gt repeat last procedure but now type
  as
• condition Exp_Resp Choice, (i.e. not
  equal)
• click Continue,
• in field Numeric
  Expression type 2,
• then OK.

5

• --gt FIRST decide Exclude Cases?
• Use Select If to exclude Practice Trials and
  (optional) trials where response was incorrect
• --gt Data, Select Cases
• Tick If condition is satisfied
• Click on IF, type/drag accuracy1
  trialtyp1, click continue
• OK

6

• do stats in SPSS --gt Exclude Cases?
  Contd
• Note when you open the file in a new session,
  you need to apply the filter again (if you so
  wish)
• --gt Data, Select Cases,
• tick Select Filter Variable, and use arrow to
  give filter variable.

7

• create subjectmeans per condition
• Example each subject has 3 responses per
• condition
• Aim for each subject, calculate the mean of the
  three responses in each condition subjectmeans
  per condition
• ---gt HOW you do this depends on the LAYOUT of
  your raw data results file (imported from
  Authorware), which obviously is different
  depending on your DESIGN!

8

• CREATE SUBJECTMEANS --gt Example 2 x 2
  repeated-measures AOV
• REQUIRED LAYOUT of datafile for AOV
• w_short w_long nw_short nw_long
• .57 .60 .64 .49
• .70 .65 .94 .80
• .56 .67 .80 .90
• etc.
• gt Factors Stimulus Type
• Length

9

• create subjectmeans for 2 x 2 within design
• ---gt Menu Analyze, Custom Tables, Basic Tables
• Summary variable RT
• Subgroup Down Subjects
• Subgroup Across Sti_Type and Length
• tick Button All combinations nested
• Statistics select Mean (Count, if you want to
  run AOV with Accuracy as DV)
• OK
• --gt print out Table from Output window and EXPORT

10

• Export table with subjectmeans import as new
  data file
• In OP window, click on table go to
• Menu File, Export, type filename filetype
  text check Options produce tab-separated
  output.
• In Notepad get rid of the top lines of table
  with text.
• Import into SPSS

11

• Import table with subjectmeans into SPSS
• --gt File Read Text Data,
• predefined format No
• delimited Yes
• variable names included No
• 1. Case begins on which line number 1
• how are cases represented Each line
• how many cases do you want to importAll
• which delimiter Tab
• Step 5 name your variables!

12

• Import table with subjectmeans into SPSS
  contd
• in Step 5 for 1. Variable (subjects) and 2.
  Variable (empty column) select Do Not Import
  from Data format - option
• Save your new data file run repeated-measures
  AOV on it.

13

• CREATE SUBJECTMEANS --gt Example 2 x 2
  between-subjects AOV
• REQUIRED LAYOUT of datafile for AOV
• Presentation Time Category Score
• 1 1
  490
• 1 1
  800
• 1 1
  900
• 1 2
  550
• 1 2
  777
• etc.
• gt Factors Presentation Time Category
  

14

• create subjectmeans for 2 x 2 between design
• ---gt Menu Analyze, Custom Tables, Basic Tables
• Summary variable RT
• Subgroup Down Subjects, Pres_Time and Category
• tick Button All combinations nested
• Statistics select Mean (Count, if you want to
  run AOV with Accuracy as DV)
• OK
• --gt print out Table from Output window

15

• Export table with subjectmeans import as new
  data file
• To export
• In OP window, click on table go to
• Menu File, Export, type filename filetype
  text check Options produce tab-separated
  output.
• To import into SPSS --gt File Read Text Data,
• etc.,
  as before.
• Save your new data file run between-subjects
  AOV on it.

16

• CREATE SUBJECTMEANS --gt Example 2 x 2 mixed
  design
• REQUIRED LAYOUT of datafile for AOV
• Gender word_score nw_score
• 1 490 500
• 1 800 870
• 1 900 930
• 1 550 700
• 1 777 888
  etc.
• gt Factors between Gender within Sti_type

17

• create subjectmeans for 2 x 2 mixed design
• ---gt Menu Analyze, Custom Tables, Basic Tables
• Summary variable RT
• Subgroup Down Subjects, Gender
• Subgroup Across Sti_Type
• tick Button All combinations nested
• Statistics select Mean (Count, if you want to
  run AOV with Accuracy as DV)
• OK
• --gt print out Table from Output window

18

• Export table with subjectmeans import as new
  data file
• To export
• In OP window, click on table go to
• Menu File, Export, type filename filetype
  text check Options produce tab-separated
  output.
• To import into SPSS --gt File Read Text Data,
• etc.,
  as before.
• Save your new data file run mixed AOV on it.

19

• do stats in SPSS run AOV
• DETAILED NOTES ON RUNNING VARIOUS AOVS
• CAN BE FOUND IN KEVINS QUANTITATIVE STATS
• LECTURES 5 AND 6
• Between-subjects AOV
• http//ibs.derby.ac.uk/kpat/QuantitativeRM/RM3_Fa
  ct_ANOVA.pdf
• Repeated-measures AOV Mixed AOV
• http//ibs.derby.ac.uk/kpat/QuantitativeRM/RM_rep
  eat_ANOVA.pdf

20

• This week Guidelines for Poster
• should be around 500 words
• for a nice sample poster see Garys Memory
  Perception Module
• http//ibs.derby.ac.uk/gary/Mpa/Sample-poste
  r/index.htm
• go to TL module web page, click on Assessment
  and check the links in the Poster section for
• --gt our Poster Feedback Sheet
• --gt a fun guide to producing poster reports

21

• Aims for next workshop
• produce poster
• think about questions for Revision Session