Lets do some testing

Difference in change between Intellectual and Social appeal for Whole cricket

A Mann-Whitney U test was used to test the hypothesis that there will be a significant difference between the change in self-reported likelihood of eating a cricket before and after by participants exposed to the intellectual appeal and the social appeal. The intellectual appeal condition had a mean rank of 181.32 and mean change of 0.23 (SD = 1.06), which was slightly greater than the social appeal condition which had the mean rank of 172.92 and mean change of 0.18 (SD = 0.82), however the intellectual appeal condition and the social appeal condition did not differ significantly (Z = -1, p = .316).

Note : Group A is the intellectual Group and Group B is the social group

# Whitney-Mann U (also known as two-sample Wilcoxon)
x1 <- dat$CricketChange[dat$Group=="Group A"]
y1 <- dat$CricketChange[dat$Group=="Group B"]
wilcox.test(x1, y1, alternative = "t", paired=FALSE)
## 
##  Wilcoxon rank sum test with continuity correction
## 
## data:  x1 and y1
## W = 15873.5, p-value = 0.3167
## alternative hypothesis: true location shift is not equal to 0
describe(dat$CricketChange[dat$Group=="Group A"])
##   vars   n mean   sd median trimmed mad min max range skew kurtosis   se
## 1    1 150 0.23 1.06      0    0.15   0  -5   6    11 0.19     13.1 0.09
describe(dat$CricketChange[dat$Group=="Group B"])
##   vars   n mean   sd median trimmed mad min max range skew kurtosis   se
## 1    1 202 0.18 0.82      0    0.09   0  -3   5     8 2.02    10.87 0.06

Difference in change between Intellectual and Social appeal for Cricket bar

A Mann-Whitney U test was used to test the hypothesis that there will be a significant difference between the change in self-reported likelihood of eating a cricket based bar before and after given by participants exposed to the intellectual appeal and the social appeal. Change was significantly larger in the social appeal condition (mean rank= 185.12, mean change = 0.63, SD = 1.26) than the intellectual appeal condition (mean rank= 164.78, mean change=0.27, SD = 1.22), Z = -2.19, p = .028.

# Whitney-Mann U (also known as two-sample Wilcoxon)
x2 <- dat$BarChange[dat$Group=="Group A"]
y2 <- dat$BarChange[dat$Group=="Group B"]
wilcox.test(x2, y2, alternative = "t", paired=FALSE)
## 
##  Wilcoxon rank sum test with continuity correction
## 
## data:  x2 and y2
## W = 13409, p-value = 0.02839
## alternative hypothesis: true location shift is not equal to 0
describe(dat$BarChange[dat$Group=="Group A"])
##   vars   n mean   sd median trimmed mad min max range skew kurtosis  se
## 1    1 150 0.27 1.22      0    0.16   0  -5   5    10 0.19     5.87 0.1
describe(dat$BarChange[dat$Group=="Group B"])
##   vars   n mean   sd median trimmed mad min max range skew kurtosis   se
## 1    1 202 0.63 1.26      0    0.42   0  -2   7     9 1.72      3.8 0.09

Correlation in age and pre-influence rating for Whole cricket and Cricket bar

A Kendall’s tau b correlation found that there was a significant weak negative association between age and initial rating likelihood of eating a cricket (tau b(352) = -0.1, p < .001, two-tailed) and likelihood of eating a cricket based bar (tau b(352) = -0.11, p = .005, two-tailed).

# Kendalls Tau between Age and BeforeCricket, 
cor.test(dat$Age, dat$BeforeCricket, method="kendall", alternative="two.sided") 
## 
##  Kendall's rank correlation tau
## 
## data:  dat$Age and dat$BeforeCricket
## z = -2.5764, p-value = 0.009985
## alternative hypothesis: true tau is not equal to 0
## sample estimates:
##        tau 
## -0.0985733
cor.test(dat$Age, dat$BeforeCricketBar, method="kendall", alternative="two.sided") 
## 
##  Kendall's rank correlation tau
## 
## data:  dat$Age and dat$BeforeCricketBar
## z = -2.7934, p-value = 0.005216
## alternative hypothesis: true tau is not equal to 0
## sample estimates:
##        tau 
## -0.1076127

Difference in gender on pre-influence rating for Whole cricket

A Mann-Whitney U test showed a significant difference between males (Mean Rank = 206.8) and females (Mean Rank= 156.25) on the self-reported likelihood of eating a cricket before appeal Z = -4.613, p = <.001.

# Whitney-Mann U (also known as two-sample Wilcoxon)
x3 <- dat$BeforeCricket[dat$Sex=="Female"]
y3 <- dat$BeforeCricket[dat$Sex=="Male"]
wilcox.test(x3, y3, alternative = "t")
## 
##  Wilcoxon rank sum test with continuity correction
## 
## data:  x3 and y3
## W = 10602.5, p-value = 3.978e-06
## alternative hypothesis: true location shift is not equal to 0
describe(dat$BeforeCricket[dat$Sex=="Female"])
##   vars   n mean   sd median trimmed  mad min max range skew kurtosis   se
## 1    1 211 4.76 3.35      5    4.58 4.45   1  10     9 0.27    -1.43 0.23
describe(dat$BeforeCricket[dat$Sex=="Male"])
##   vars   n mean   sd median trimmed  mad min max range  skew kurtosis   se
## 1    1 141 6.45 3.11      7    6.68 4.45   1  10     9 -0.48    -1.13 0.26

Difference in gender on pre-influence rating for Cricket bar

A Mann-Whitney U test was used to test the hypothesis that there was a significant difference between males and females on the self-reported likelihood of eating a cricket bar before appeal, and revealed that the male (Mean Rank = 199.10) and female (Mean Rank=161.4) groups did differ significantly (Z = -3.46 p = .001)

x4 <- dat$BeforeCricketBar[dat$Sex=="Female"]
y4 <- dat$BeforeCricketBar[dat$Sex=="Male"]
wilcox.test(x4, y4, alternative = "t")
## 
##  Wilcoxon rank sum test with continuity correction
## 
## data:  x4 and y4
## W = 11688.5, p-value = 0.000534
## alternative hypothesis: true location shift is not equal to 0
describe(dat$BeforeCricketBar[dat$Sex=="Female"])
##   vars   n mean   sd median trimmed  mad min max range  skew kurtosis   se
## 1    1 211 6.34 3.24      7    6.55 4.45   1  10     9 -0.34    -1.33 0.22
describe(dat$BeforeCricketBar[dat$Sex=="Male"])
##   vars   n mean   sd median trimmed  mad min max range skew kurtosis   se
## 1    1 141 7.63 2.63      9    8.03 1.48   1  10     9   -1    -0.06 0.22

Back to the main page

Main page