Date post: | 30-Dec-2015 |
Category: |
Documents |
Upload: | megan-lucas |
View: | 112 times |
Download: | 3 times |
RISKY SHIFT:
LAB REPORT WRITE-UP
Week 7 Practical
WEEK 1
WEEK 2
WEEK 3
WEEK 4
WEEK 5
WEEK 6
WEEK 7
WEEK 8
WEEK 9
WEEK 10
LECTURE PRACTICAL
NONPARAMETRICS 1 1ST PRACTICAL
NONPARAMETRICS 2 1ST ANALYSIS
SAMPLING DISTRIBUTIONS
1ST ANALYSIS + PROBLEMS 1
HYPOTHESIS TESTING
2ND PRACTICAL
RELATED T-TEST
2ND ANALYSIS + SOLUTIONS 1
INDEPENDENT T-TEST
INDEPENDENT ANOVA
DEPENDENT ANOVA
2ND ANALYSIS + PROBLEMS 2
3RD PRACTICAL
3RD ANALYSIS + SOLUTIONS 2
3RD ANALYSIS + PROBLEMS 3
NO LECTURE NO PRACTICAL
WEEK 7 PRACTICALRISKY SHIFT
Tackle an additional problem sheet which should help you understand the principles behind non-parametric tests more fully.
Learn to calculate the Wilcoxon test by hand.
Understand the way in which the practical reports will be marked and what you have to do to do well!
BY THE END OF THE SESSION, YOU SHOULD BE ABLE TO:
Clarify any additional problems that you may have regarding this lab report with your friendly neighbourhood demonstrators.
LEARNING OUTCOMESRISKY SHIFT
PRACTICAL WRITE-UP
INTRODUCTION
METHOD
RESULTS
DISCUSSION
ABSTRACT
TITLE
REFERENCES
APPENDICIES
212
20
20
20
20
6
0
IV and DV included, precise, concise
Maximum 150 words, summary of all sections
Clear structure leading to experimental rationale
Evidence of extra reading, beyond the handouts
Sufficient information to replicate the study
Participants, materials, design, procedure
Clear and logical descriptive and inferential stats
Well designed figures related to the findings
Summary of main results, relating to introduction
Evaluation of study, suggestions for future
APA formatted, consistent with references in text
Handouts, SPSS printouts, workings out…
RISKY SHIFT
WILCOXON BY HANDRISKY SHIFT
DOWNLOAD AND SAVE THE FILE ‘WILCOXONBYHAND’
First, you should be able to complete the smaller exercise on the Wilcoxonbyhand document, and check
your workings using ‘wilcoxonbyhand_done’
There is an addition exercise which provides you with the steps to carry out a Wilcoxon by hand on the Risky
Shift data using Excel. Taking a look at this exercise should improve your Excel skills if nothing else.
DO NOT PANIC ABOUT GETTING THROUGH BOTH EXERCISES.
FOCUS ON UNDERSTANDING HOW TO DO WILCOXON BY HAND.
DOWNLOAD AND SAVE THE FILE ‘WILCOXONBYHAND_DONE’
WILCOXON BY HANDRISKY SHIFT
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
NON-PARAMETRIC TESTS (1)
GROUP 1: Radiohead
GROUP 2: LCD Soundsystem
03, 05, 04, 04, 07, 10, 04, 05, 05
13, 17, 12, 17, 06, 10, 16, 17, 19
(a) Find the difference between score pair and retain the sign
(b) Rank the differences ignoring their sign. Ignore zero differences
(c) Ranking criteria as in Mann-Whitney
WILCOXON TEST
03, 05, 04, 04, 07, 10, 04, 05, 05
13, 17, 12, 17, 06, 10, 16, 17, 19
10, 12, 08, 13, -01, 00, 12, 12, 14
(-)01, 08, 10, 12, 12, 12, 13, 14
(-)01, 02, 03, 04, 05, 06, 07, 08
(-)01, 02, 03, 05, 05, 05, 07, 08
NON-PARAMETRIC TESTS (1) WILCOXON TEST
GROUP 1: Radiohead
GROUP 2: LCD Soundsystem
03, 05, 04, 04, 07, 10, 04, 05, 05
13, 17, 12, 17, 06, 10, 16, 17, 19
(-)01, 02, 03, 05, 05, 05, 07, 08
(d) Add positive and negative signed ranked independently
(e) Note the smaller sum of ranks (W)
(f) Note the number of non-0 differences (N)
Negative Ranks = 01 Positive Ranks = 35
W = 1
N = 8
Here’s a reminder of how to do the Wilcoxon ranking from the lecture…
Follow the sheet to complete a Wilcoxon for the small data set.
Tackle an additional problem sheet which should help you more fully understand the principles behind non-parametric tests.
Learn to calculate the Wilcoxon test by hand.
Understand the way in which the practical reports will be marked and what you have to do to do well!
BY THE END OF THE SESSION, YOU SHOULD BE ABLE TO:
Clarify any additional problems that you may have regarding this lab report with your friendly neighbourhood demonstrators.
LEARNING OUTCOMESRISKY SHIFT
YOUR CHOICE
YOU HAVE THE REST OF THE SESSION TO FOCUS ON WHAT YOU THINK YOU NEED TO IMPROVE YOUR UNDERSTANDING
LAB REPORT TESTS BY HAND NONPARAMETRICS
Try working through the second Wilcoxon
analysis using the ‘riskyinexcel’ workbook and
compare your hand results with the
results you obtained from SPSS.
The brave could also try a Friedman’s test
by hand using Graham’s guide.
Make sure you are happy with your
interpretation of the SPSS output in
relation to the three experimental
hypotheses ,and make sure you know what
needs to go into your lab report to get the
best marks.
Download ‘problem5’ and start to work
through the questions related to the
forthcoming related t-test (as well as some problems relating to
correlation and regression).
RISKY SHIFT
DOWNLOAD AND SAVE THE FILE ‘RISKYINEXCEL’
WILCOXON BY HANDRISKY SHIFT
There’s an awful lot of data here, so it’s probably best to focus on one of the hypotheses and take it slowly.
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
At the bottom of the sheet (and the
separate Word file), there are instructions on how to calculate
Wilcoxon using Excel.
DOWNLOAD AND SAVE THE FILE ‘WILCOXONBYHAND’
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
1) CALCULATE THE DIFFERENCE BETWEEN D AND C SCORES IN (H) AND RETAIN THE SIGN. = CELLab- CELLcd
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
2) CALCULATE THE ABSOLUTE DIFFERENCE FOR (H) IN (J) =ABS (CELL)
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
3) COPY AND PASTE ID AND SCORE VALUES FROM COLUMN I AND J INTO COLUMN K AND L EDIT > COPY
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
4) SORT COLUMNS K AND L BY COLUMN L (RANKED DIFFERENCES) DATA > SORT > SORT BY
We are doing this as an easy way to rank order
the data, but we also need to keep track of whether
these differences are positive or negative in the
final analysis. Hence, each individual difference keeps its unique id code.
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
5) BEGINNING WITH THE FIRST NON-ZERO DIFFERENCE IN COLUMN N ASSIGN RANK SCORES STARTING AT 1
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
6) SCORES THAT SHARE THE SAME NON-ZERO DIFFERENCE MUST SHARE AVERAGE RANK IN COLUMN P
The first 11 numbers all share a difference of 0.4,
so they are all assigned the rank
value of 6 (1+2+3+4+5+6+7+8
+9+10+11/11 = 6)
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
7) TAKE THE SORTED ID VALUES FROM COLUMN K AND PASTE INTO COLUMN O
8) NOTE THAT IN COLUMN S, THE ORIGINAL SIGNED DIFFERENCES ARE ALREADY THERE
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
9) SORT COLUMNS O AND P BY COLUMN O (ASCENDING) DATA > SORT > SORT BY
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
10) COPY AND PASTE VALUES FROM COLUMN P OVER TO COLUMN U DATA > SORT > SORT BY
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
11) REINSTATE NEGATIVE VALUES WHERE APPROPRIATE
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
12) SORT COLUMN U (ASCENDING)
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
Congratulations! You now have everything you need to finish the test!
SUM NEGATIVE SUM POSITIVE
No. NEGATIVE No. POSITIVE TOTAL OBSERVATIONS
828 33
37 4 41
Ranks
37a 22.38 828.00
4b 8.25 33.00
7c
48
Negative Ranks
Positive Ranks
Ties
Total
grp mean1_5 -pre mean1_5
N Mean Rank Sum of Ranks
grp mean1_5 < pre mean1_5a.
grp mean1_5 > pre mean1_5b.
grp mean1_5 = pre mean1_5c.
NOTE: This is exactly the same data
produced by SPSS!
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
Congratulations! You now have everything you need to finish the test!
SUM NEGATIVE SUM POSITIVE
No. NEGATIVE No. POSITIVE TOTAL OBSERVATIONS
828 33
37 4 41
We use the smaller of the sums (33) as our observed W. We use the total number of non-zero observations (41) as our df.
Look up the critical value of W, taking into account N. If our obtained W is smaller than the critical value of W, we reject the null
hypothesis and conclude that two groups do differ significantly.
WILCOXON BY HANDRISKY SHIFT
Q2: Is the average level of risk recorded in the PRE assessment different from the risk recorded in the GROUP assessment?
Wilcoxon testRATINGS + WITHIN Ss + 2 GROUPS
Obtained W = 33
Critical W = 89
N = 41
Obtained W < Critical W
There is a statistically significant difference
between the two groups
Table only goes up to N = 25 but this is good enough (the critical value would be much bigger than 89 for an N = 41)