Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Multiple Comparison Procedure For Use With Friedman Test ( Comparing All Treatments With A Control )
5/5/12
Learning Outcome
At the end of this lesson student should be able to:
Compute all treatments with a control and compare of two or more treatments, of which one is control condition by using Friedmens Test.
5/5/12
Assumptions
The
data consist of b mutually independent samples (blocks) of size k, the number of treatments, excluding the control condition. by xi0 and xij (i = 1, ,b and j = 1,,k) the responses to the control and the jth treatment in the ith block of a randomized complete block design. 5/5/12
Represent
Hypotheses
Case A Case B Case C
H0 : Mj M0 H0 : Mj M0 H0 : Mj = M0 H1 : Mj < M0 H1 : Mj > M0 H1 : Mj M0
5/5/12
t
Block 1 2 : : b
Test Statistic
Treatment 0 X10 X20 : : Xb0 1 X11 X21 : : xb1 k X1k X2k : : Xbk
Signed differences ,dij = xij - xi0 Number of minuses sign, rj Number of pluses sign,b - rj
5/5/12
Continue
Decision Rule
Case A
5/5/12
Back
5/5/12
CASE
HYPOTHESIS
DECISION
H0 : Mj M0 H1 : Mj < M0
H0 : Mj M0 H1 : Mj > M0
H0 : Mj = M0 H1 : Mj5/5/12 M0
Example
A paint manufacturer wished to compare three newly proposed with the standard formula currently in use for manufacturing a particular type of house paint. The response variable was a composite score that incorporated several properties that measure paint quality. Since the effective quality of paint is affected by the surface to which it is applied, 5/5/12 the company researches applied
0( Stand ard) 13 12 15 14 13
1 25 27 29 21 31
2 17 15 19 9 27
3 25 25 23 13 21
Solution
Hypotheses: A high score indicates a higher
quality
0( Standa 1 2 3 rd) 13 25(+12) 17(+4) 25(+15 ) 12 27(+15) 15(+3) 25(+13 ) 15 29(+14) 19(+4) 23(+8) 14 13 21(+7) 9(-5) 13(-1) 31(+18) 27(+14 21(+8) ) 0 1 1 5 4 4
An experimentwise error rate, = 0.10 Refer table A.15 for k = 3 (treatment) b = 5 (surfaces) Critical Value = 0
5/5/12
5/5/12
Decision:
Formula
Conclusion:
Formula
There is enough evidence to support the claim that a high score indicates higher quality.
Formula
There is not enough evidence to support the5/5/12 claim that a high score
EXERCISES
5/5/12
Exercise 1
Syme and Pollard* conducted an experiment to investigate the effect of different motivation levels on measures of food-getting dominance in the laboratory rat. The data shown in Table 7.10 are the amounts of food in grams eaten by eight male hooded rats following 0 as control condition 24 5/5/12 and 72 hours of food deprivation.
Table 7.10 : Amount of food, grams, eaten by eight rats under three levels of food deprivation Hours of food deprivation Subject 0 24 72 1 2 3 4 5 6 7 8
4.5 15.3 9.6 7.2 6.8 4.7 10.1 9.3
5/5/12
Exercise 2
A paint manufacturer wished to compare three newly proposed with the standard formula currently in use for manufacturing a particular type of house paint. The response variable was a composite score that incorporated several properties that measure paint quality. Since the effective quality of paint is affected 5/5/12
Table 7.13
Surface A B C D E
Formula 0 13 12 15 14 13
5/5/12
1 25 27 29 21 31
2 12 10 13 9 11
3 11 9 12 13 10
Exercise 3
Syme and Pollard* conducted an experiment to investigate the effect of different motivation levels on measures of food-getting dominance in the laboratory rat. The data shown in Table 7.10 are the amounts of food in grams eaten by eight male hooded rats following 0 as control condition 24, 72 and 96 hours of food deprivation. A higher amount of food indicates more effects to the three levels of food deprivation. Let = 0.05
*Syme, G. J., and J. S. Pollard, The Relation between Differences in 5/5/12 Level of Food Deprivation and Dominance in Food Getting in the Rat,
Table 7.10 : Amount of food, grams, eaten by eight rats under three levels of food deprivation Hours of food deprivation Subject 0 24 72 96 1 2 3 4 5 6 7 8 4.5 15.3 9.6 7.2 6.8 4.7 10.1 9.3 5.9 8.1 8.1 8.6 8.1 5.9 9.5 7.9 13.9 12.6 8.1 6.8 14.3 4.2 14.5 7.9 5.6 19.8 15.3 14.4 8.9 10.1 12.5 14.7
5/5/12
Link to youtube :
http://www.youtube.com/watch?v=gI0
5/5/12
Thank you..
5/5/12