Assessment Two tests your ability to conduct a factorial ANOVA and make appropriate decisions regarding any follow up comparisons. The following data set is adapted from Jaccard and Becker (2002). It examines different ways in which information is communicated from one individual to another. Three of the most common ways are by email, telephone, and written notes or letters.
In order to determine whether the way in which information is communicated influences its effectiveness, Jaccard and Becker proposed an experiment in which participants are presented the identical information using email, a telephone call, or a written note. The length of the message is also varied such that one set of messages contains only the relevant information, another set contains both the relevant information and a small number of extraneous details, and the third set contains both the relevant information and a large number of extraneous details. The design is thus a 3 x 3 factorial with three levels of message format (email, telephone, written) and three levels of message length manipulated by amount of extra details (no extras, small, large).
This is a fully randomised factorial design. After reading or listening to their assigned message, participants responded to a series of questions about the message content. The dependent variable is the number of correctly answered questions.
PART ONE: The Omnibus ANOVA
The score will then be halved to give a final score out 10 towards the total of 35.
Preliminary Inspection for Homogeneity of Variance
Are the variances statistically equivalent across groups? In the adjoining box to the right state either YES or NO
Provide statistical evidence for your decision
Analysis of Main Effects
1. Message Format
In the box to the right report the main effect for message format as
F ( , ) = , p =
F ( , ) = , p =
Is the main effect for message format statistically significant? In the adjoining box to the right state YES or NO
Give the values of the means being compared and their corresponding level of the factor
Give Level of the IV Give Mean
2. Message Length
In the box to the right report the main effect for message length as
F ( , ) = , p =
F ( , ) = , p =
Is the main effect for message length statistically significant? In the adjoining box to the right state YES or NO
Give the values of the means being compared and their corresponding level of the factor
Give Level of the IV Give Mean
Comparison of the Main Effects
Which of the main effects accounts for more of the variance in the number of correct responses? Put the name of the factor corresponding to that main effect in the box to the right
How much more variance does the factor stated above account for compared to the other factor?
Analysis of the Interaction
In the box to the right report the statistical analysis of the interaction as
F ( , ) = , p = .
F ( , ) = , p =
Is the interaction effect statistically significant? In the adjoining box to the right state YES or NO
What does the test of the interaction tell us about the effects of the two factors (message format and length). Please type your response in the box below. Note that I have left four lines. A good answer can be given in less than four lines.
PART TWO: Follow Up Analyses
Simple Effects Analysis
For each of the simple effects that you think should be conducted, you are asked to fill in the information asked for in the box below in italics and found in the square brackets.
Further analyses examined the simple effect for [state the IV]2 marks when [state which level of which IV is being tested]2 marks. Statistical significance was evaluated at an alpha level of [state the alpha level]2 marks. The simple effect was found to be statistically [state whether significant or non significant]2 marks, [report the statistical result as F( , ) = , p = ]1 mark. Further analysis with simple comparisons were [state whether conducted/ not conducted]1 mark.
Instruction: delete the square brackets and replace it with the requested information. For example, if your factor was gender:
Further analyses examined the simple effect for [state the IV]2 marks when...etc
Would be changed to:
Further analyses examined the simple effect for gender when...... etc.
Cut and paste the answer box and fill it in for each simple effect you decide to report. For example, if you had 3 simple effects to report you would need 3 answer boxes.
How a mark out of 10 for the simple effects will be calculated: each answer box is worth 10 marks. The average of the answer boxes will be calculated. Note: the answer box will get a mark of 0 if it is inappropriate. In addition, if a simple effect is missing then a mark of 0 will be given and treated as if it an answer box had been provided.
Simple Comparisons Analysis
Provide the information below for each of the simple comparisons you would conduct.
Simple comparisons were evaluated at an alpha level of [state alpha level]2 mark. The difference between the mean for [state condition and (M = )]2 mark and the mean for [state condition and (M = )] 2 marks was found to be statistically [state whether significant/non significant]2 marks, [report the statistical result using the format F( , ) = , p = ]
Cut and paste the answer box and fill it in for each simple comparison you decide to report. For example, if you had 3 simple comparisons to report you would need 3 answer boxes.
How a final mark out of 8 for the simple comparisons will be calculated: each answer box is worth 8 marks. The average of the answer boxes will be calculated. Note: the answer box will get a mark of 0 if the simple comparison is inappropriate. In addition, if a simple comparison is missing then a mark of 0 will be given and treated as if it an answer box had been provided.
Giving a Reason why you have reported Simple Effects and Comparisons
Why are simple effects and simple comparisons appropriate here? Note that I have left four lines. A good answer can be given in less than four lines.
PART THREE: Interpretation of the results
What conclusions can you draw about the effects of format and length of a message on the effectiveness of a message? I want you to put the results together to draw informative conclusions about the effectiveness of the different types of message. Include in your discussion practical recommendations from this research.
You do not need any additional information about the design of the study in order to answer this part.