Analysis
ANOVA / MANOVA
A two-way analysis of variance (ANOVA) and a multivariate analysis of variance (MANOVA) were conducted for the brand ‘Nature Valley’ cereal bar. The purpose of the analysis is to test for any significant relationship between the independent variables and the dependent variables. For the ANOVA test the independent variables are gender and their pre to post change score. The dependent variable is the Likert item “Nature Valley has better health benefits than other bars.” For the MANOVA test the same independent variables, gender and change score, are used with the dependent variables being all 10 Likert items. There were 63 respondents in this survey.
Table 1. Descriptive Statistics (ANOVA)
|
Group Mean Scores and Standard Deviation |
Move-Up |
Same |
Move-Down |
Male |
Mean: 3.9
Standard Deviation: .9
N=9 |
Mean: 3.6
Standard Deviation: .5
N=10 |
Mean: 2.9
Standard Deviation: .6
N=9 |
Female |
Mean: 2.7
Standard Deviation: .6
N=3 |
Mean: 3.2
Standard Deviation: 1.1
N=5 |
Mean: 3.1
Standard Deviation: .6
N=27 |
(N=63)
Table 1 reflects descriptive statistics for one Likert item (Nature Valley has better health benefits than other bars.) and pre-post exposure change scores for both males and females. Each cell shows the average change score, standard deviation, and sample size for Up-Movers and Down-Movers, and for those respondents whose degree of congruence remained the same post ad exposure. Men have higher mean scores than women except for Down-Movers and Up-Movers have the highest score. The greatest difference between mean scores is between male and female Up-Movers. This conveys that, after viewing the ad, men are more likely than women to consider Nature Valley has better health benefits than other bars. The majority of both males and females are Down-Movers suggesting that most respondents did not agree with the statement, "Nature Valley has better health benefits than other bars." after seeing the ad. Interestingly, an overwhelming 78 percent of female respondents exhibited negative change scores.
Table 2. ANOVA
|
Sum of Squares |
Degrees of Freedom |
Mean Square |
F-ratio |
Change score |
1.58 |
2 |
.79 |
1.72 |
Gender |
2.19 |
1 |
2.19 |
4.76* |
Change score by gender |
3.69 |
2 |
1.84 |
3.99* |
*p ≤ .15
The F-ratios for gender and change score by gender (interaction) are significant and therefore these results can be projected to a larger population. Thus, in 85 or more samples drawn from the same population as this sample, it would be expected that the result would be about the same as this sample.
Table 3. Descriptive Statistics (MANOVA)
Two Way Factorial Multivariate Analysis of Variance (N=63) |
|
Male (N=28) |
Female (N=35) |
Up
(N=9) |
Same (N=10) |
Down (N=9) |
Up
(N=3) |
Same
(N=5) |
Down (N=27) |
|
Mean |
Standard Deviation |
Mean |
Standard Deviation |
Mean |
Standard Deviation |
Mean |
Standard Deviation |
Mean |
Standard Deviation |
Mean |
Standard Deviation |
Good brand |
4.0 |
.5 |
3.9 |
.6 |
3.9 |
.6 |
3.7 |
.6 |
4.0 |
.7 |
3.7 |
.7 |
Energy |
3.6 |
.7 |
3.5 |
.5 |
3.3 |
.7 |
3.7 |
.6 |
4.0 |
.7 |
4.4 |
.6 |
Healthy |
3.9 |
.9 |
3.6 |
.5 |
2.9 |
.6 |
2.7 |
.6 |
3.2 |
1.1 |
3.1 |
.6 |
Substitute |
3.3 |
.9 |
3.0 |
1.0 |
2.4 |
1.1 |
3.3 |
1.2 |
3.2 |
1.3 |
3.3 |
.9 |
Expensive |
2.9 |
.8 |
3.4 |
.5 |
2.6 |
.7 |
3.3 |
.6 |
3.2 |
.8 |
3.1 |
.5 |
Appetite |
3.3 |
.7 |
3.3 |
1.1 |
3.0 |
.7 |
3.0 |
1.0 |
3.8 |
.4 |
3.2 |
.8 |
Good Taste |
3.1 |
1.1 |
3.5 |
1.0 |
3.1 |
.6 |
3.7 |
.6 |
3.4 |
.9 |
3.1 |
.9 |
Purchase |
3.8 |
.4 |
3.7 |
.8 |
3.6 |
.5 |
3.7 |
.6 |
3.6 |
.5 |
3.3 |
.8 |
Recommendation |
3.7 |
.7 |
2.9 |
.9 |
3.1 |
.9 |
3.3 |
.6 |
3.6 |
.5 |
3.3 |
.8 |
Attractive |
4.1 |
.6 |
3.7 |
.5 |
3.0 |
.9 |
4.0 |
1.0 |
3.2 |
.8 |
3.0. |
.8 |
Table 3 reflects descriptive statistics for pre-post exposure change scores for males and females in relation to all ten Likert items. Among Up-Movers, there were notable differences between the mean scores of males and females on two of the ten brand attributes. More men believed Nature Valley has better health benefits than other bars while more women believed Nature Valley is not a good tasting cereal bar (reverse scale). Among those respondents whose change scores remained the same, there was one notable difference between males and females; females more likely to believe that they feel good about purchasing Nature Valley. Among down-movers, there were two notable differences among male and female; females are more likely to believe that Nature Valley makes them feel energized and they would eat Nature Valley cereal bars as a substitute for a meal.
Table 4. MANOVA
|
Wilk’s Lambda |
F-ratio |
Degrees of Freedom |
Change score |
.61 |
1.37 |
20 |
Gender |
.78 |
1.32 |
10 |
Change score by gender |
.64 |
1.19 |
20 |
*p ≤ .15
When comparing the ten brand attributes to the independent variables, none of the F-ratios were found to be significant. Thus, in 85 or more samples drawn from the same population as this sample, it would be expected that the result would be about the same as this sample.