Questions tagged [contrasts]
In linear models and particularly in ANOVA, a contrast is a linear combination of parameters with coefficients summing up to zero. It is used to test the corresponding null hypothesis. Contrasts are especially often used with categorical predictors (factors) to make comparisons among the groups (categories). [See also tag 'categorical-encoding']
443
questions
0
votes
0
answers
17
views
Effects Contrast Matrix for Interaction with Multiple Grand Means
I want to include an interaction between two factors: area and house_type, in my GAM (MGCV), I hope to make a sum-to-zero ...
- 76
0
votes
0
answers
23
views
The problem of double adjustment of p-values in multilevel models with categorical predictors
I have a set of around 30 multilevel models which have the same predictor variables (and the same cluster variable), but different outcome variables. Since there are so many models, I want to perform ...
- 357
4
votes
2
answers
67
views
Making 3 comparisons with a ternary variable
I have a study in which I want to study the effect of a ternary variable $x$ in interaction with a binary variable $y$. The ternary variable is categorical, and for some reasons, I want to do the $3$ ...
5
votes
1
answer
69
views
ANOVA Type III understanding
I stumbled across something very interesting to me earlier, in the question and answer here:
ANOVA type III strange results
I reiterate the warning I found in the function documentation, highlighting ...
- 305
0
votes
1
answer
46
views
Emtrends pairwise comparisons all have same standard errors. 3 way interaction
I have created a linear mixed model and am using emtrends to determine the impact of a dichotomous variable on the effect of a continuous variable, across levels of a categorical variable. Is it ...
- 11
1
vote
0
answers
31
views
Constructing subgroup comparisons for variable selection
I am trying to construct a covariate according to the the following description:
When constructing the covariates to measure a subgroup effect, we
generate covariates that capture the causal effect ...
- 63
0
votes
0
answers
23
views
Is the intercept of a complex sum-coded regression basically useless for interpretation? Maybe even for some simple models?
In regression analysis, one may choose to code categorical variables differently depending on interpretability considerations. One such coding scheme known as sum coding (a kind of effect coding ...
- 272
0
votes
0
answers
23
views
What is the name for contrast coefficients where the difference between the coefficients = 1?
I am wondering if there is a name for the type of contrast coding where the difference between the terms is equal to 1. Let me illustrate with some fake data (using R code, apologies)
Two groups, with ...
- 2,043
0
votes
0
answers
14
views
How to translate the set of contrasts over model coefficients into definitions of two nested models for Likelihood Ratio testing?
With the data as below:
Categorical predictor: "Group" with 2 levels: Group 1 and Group 2
Categorical predictor: "Treatment" with 3 levels: A, B, C
Categorical binary response: &...
- 11
1
vote
0
answers
34
views
How to test specific contrasts about levels of categorical variables through nested models? [closed]
This is not about obtaining any dataset. I HAVE the dataset.
This is not about debugging code, this is about EXPLAINING the way to obtain statistical relationship between the nested models (LRT ANOVA) ...
- 11
0
votes
0
answers
38
views
Multilevel-Regression Contrasts and Multiple Comparisons
My experiment looks roughly something like this:
5 Different biological replicates were tested with 4 Conditions (incl. Control) and 3 Time points (1h, 6h, 24h). My samples are nested within ...
0
votes
0
answers
22
views
Contrasts by each categorical factor or by all?
I have the following model:
Cmic_mg_cm2 ~ group*sample_date + mean_ph + mean_shan_veg +
mean_tree_shandiv + (1|plot)
with both group and sample_date ...
- 31
0
votes
0
answers
10
views
Why do we use n-1 effect codes and -1 values in sum (effect) coding in linear regression? Why can't we treat all the effects equally? [duplicate]
When doing effect/sum coding in linear regression (which AFAIK are the same), the contrasts are coded as:
condition1
condition2
condition3
condition1
1
0
0
condition2
0
1
0
condition3
0
0
1
...
- 1
1
vote
1
answer
25
views
How to set contrasts for contrast in post-hoc comparison of linear mixed effect model in R?
i am new in R and studying statistical analysis for experiments, please help me~
I have 3 age groups (Y/HO/LO) and 2 conditions (Related/Unrelated). I am running a linear regression model on RT (...
- 13
0
votes
0
answers
24
views
Which of these is the correct way to encode a type III ANOVA analysis?
I am trying to run a type III ANOVA analysis to look at how the three factors A, B and C affect the continuous variable X. Online I have found two ways of encoding the model which give different ...
- 441