Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
17 Cards in this Set
- Front
- Back
|
What is meant by the terms factorial design and factor, and what are the advantages of a factorial design? |
A factorial design is one in which we have multiple grouping variables (or IVs) and we pair every level of Factor 1 with every level of Factor 2, etc. for all factors. Suppose we have a factor for gender (Male v Female) and a factor for age (adult v child) and we’re looking at time spent watching sports on tv. In a factorial design, we have four cells (or groupings) to potentially compare: men, women, male children and female children.Factor is simply another name for IV (or grouping variable). |
|
|
What are main effects and interactions, and how does the presence of an interaction influence our interpretation of a main effect? |
Main effects: A main effect is the effect of one IV ignoring levels on other IVs in the analysis. For instance, we may explore the main effect for gender (ignoring the fact that individuals can also be split into adult and child groupings).Interactions: An interaction is when the effect of one IV on the DV is different at various levels of another IV. For instance, for adults, we may observe that men watch more sports than women, but for children, we observe that male children watch less sports than female children. That is, the effect of gender for sports viewing is dependent on (or interacts with) age.The presence of an interaction effect means that we may not be able to generalize a main effect across all conditions, since it is conditional on another IV. For instance, if men watch more sport than women, but male children watch less sport than female children, then any overall effect of gender would be misleading (since the effect runs in opposite directions based on participant age). On the other hand, if the gender difference is in the same direction for children and adults (albeit, the magnitude of the difference varies), then we can still conclude that there is a meaningful gender difference overall. |
|
|
What are cell means and what information do they provide? |
A cell is a particular configuration of IV levels. For instance, with 2 levels for age and 2 for gender, we have 4 different configurations of the various IV levels. One cell may have participants who are male and adult, another contains individuals who are female and child-age, etc. Cell means are simply the means for individuals with that particular configuration of IV levels. Cell means allow us to ascertain if there are any differences among all of the possible groupings in our dataset. |
|
|
What means do we use to examine (a) main effects and (b) interaction effects? |
Main effects are computed by using the means for the various levels of a particular IV. For gender, we would have a mean score for males, and a mean score for females.Interaction effects are examined by using the cell means. |
|
|
What is a simple effect, and how do simple effects differ from main effects? |
Simple effects are the effect of one factor at one level of the other factor. For instance, we could look at the simple effect for gender in the group of adults only. Another simple effect for gender would be observed in the group of children only. Main effects, on the other hand, simply group children and adults together when looking at gender differences. |
|
|
Under what conditions would you expect the simple effects of a variable to be the same as the main effects of that variable? |
The simple effect should be consistent with the main effect when that particular IV does not significantly interact with another IV in the factorial design. |
|
|
How does the calculation of a simple effect differ from the calculation of a one-way ANOVA on the same data? |
There are more significance tests for simple effects since the main effect for the IV tested within the ANOVA context must be tested for each different level of another IV within the simple effects context. For instance, suppose that an IV has three levels (country of origin: Australia, New Zealand, and England), each with 20 participants (20 x 3 = 60 participants in total). Suppose that individuals from these various countries could be split further by gender (and for simplicity sake let’s say that this results in 6 groups of 10 = 60 participants). ANOVA ignores the fact that it could do group comparisons of all 4 groups and instead compares the means of individuals from Australia (N=20), New Zealand (N=20), and England (N = 20). |
|
|
What do the sums of squares of all the simple effects of one factor add up to in a two factor experiment? |
The sum of squares of all the simple effects of one factor add up to the sum of squares for the main effect for that factor + the sum of squares for the interaction effect. |
|
|
How do you choose which simple effects to test if there is a significant interaction? |
Only test simple effects you are interested in. For instance, if you have gender and age effects for sports watching, but you’re only really interested in the gender effects, then you may conduct simple effects for gender at the different age groupings. |
|
|
What is meant by a three-way interaction? |
A three-way interaction occurs when the interaction between two IVs differs across levels of another IV. For instance, the interaction between A and B is different at C = 1 than it is at C = 2. |
|
|
ow does the existence of a 3-way interaction influence the interpretation of a 2-way interaction and a main effect within the same analysis? |
A sig. 3-way interaction means that the 2-way interaction may vary at different levels of the third IV. It also may mean that main effects of one IV do not hold at various levels of IV2 and IV3. |
|
|
If a 3-way interaction is found, what further analyses need to be carried out? |
Simple interaction effects should be examined at each level of a third IV. |
|
|
If there is no 3-way interaction, but there is a two-way interaction in a 3-factor design, what further analyses need to be carried out? |
Simple main effects for one of the IVs at the various levels of the other IV in the 2-way interaction. |
|
|
How many different ways are there to graph a 3-way interaction and how should these be decided between? |
6.They should be decided on theoretical basis – which simple interaction effects would you like to plot separately at different levels of a third IV?In the absence of any clear rationale, you may use the following guidelines:Select the IV with the largest number of levels to be on the horizontal axisSelect the IV with the least number of levels to be shown as separate plotsSelect the remaining IV for separate lines within each of the plots. |
|
|
Plots are a useful tool to determine the nature of the effects of variables. If a plot for a2-way ANOVA showed parallel lines, this would indicate... - That one group scored consistently higher than another group, which does notrepresent an interaction - Both main effects and interactions may be significant - Simple effects should be considered - That one group scored consistently higher than another group, which represents aninteraction |
- That one group scored consistently higher than another group, which does not represent an interaction |
|
|
The simple effects of a variable will be the same as the main effect of that variablewhen: - The simple effects are not significant - The interaction is significant - There is no interaction - They (main effects, simple effects and interactions) are all significant |
There is no interaction |
|
|
In a three-way factorial ANOVA, all main effects, all two-way interactions and the three-way interaction were found to be significant. What further analysis should the researcherundertake? - None - Simple interaction effects and simple effects - Simple effects - Post hoc linear contrasts |
- Simple interaction effects and simple effects |
