Description Usage Arguments Value Author(s) References See Also
Deducer plug-in enabling the analysis of factorial ANOVA designs, including within-subjects and mixed designs. Includes functions for reshaping within-subjects data from wide to long format.
1 |
data |
A data.frame containing the variables to be analyzed. |
dv |
A dot object such as .(dv). The dependent variable. |
wid |
A dot object such as .(wid). A unique identifier for each case or subject in the dataset. Will be converted to factor (with a warning). |
between |
Dot object such as .(payment,courtesy). A vector of between-subjects factors to be included in the model. |
observed |
Dot object such as .(status). A vector of observed (not manipulated) factors to be included in the model, distinguished only for calculations of generalized eta squared (see Bakeman, 2005). |
within |
Dot object such as .(time). A vector of within-subjects factors. If data are in wide format (one row per subject), the data will need to be converted to long format, in which a new variable is added that identifies the level of the within-subjects factor for a particular observation. |
type |
Type of Sum of Squares, using the SAS sQuote(Type II) or sQuote(Type III) terminology. Type 3 is the default, and tests the effects of each factor by comparing the fit of a model with all the terms to the fit of a model with all the terms except the one being tested. To provide results comparable to those produced by SPSS and SAS, DeducerEZ stipulates options(contrasts=c(contr.sum,contr.poly)). |
detailed |
Logical. Should detailed output, including Sum of Squares for each factor, be provided? Default is FALSE. |
descriptives |
Logical. Should descriptive statistics for the model be provided? Default is FALSE. |
Tukey |
Logical. If true, Tukey's Honestly Significant Difference for all possible pairs is computed on all between-subjects factors and interactions of between-subjects factors. This is accomplished by first constructing a linear model by crossing all between-subjects factors, calling aov on the lm, and then calling TukeyHSD on the aov. |
x |
Variable to be plotted along the x-axis. If blank, no plot is generated. |
split |
Variable to be used in the legend. Can be left blank to create a plot with just one factor. |
x_lab |
Character string used for the label of the x-axis. |
y_lab |
Character string used for the label of the y-axis. |
split_lab |
Character string used for the label of the legend. |
test.var |
Dot object such as .(payment) specifying the factor levels which will be compared (in pairs) using the sme (Simple Main Effects) function. |
at.var |
Dot object such as .(courtesy). Levels of test.var will be compared AT each level of at.var. Passed to the sme function. |
var.equal |
Logical value. Are the variances of the cells in the factorial design assumed to be equal? Default is FALSE. Passed to the sme function. |
p.adjust.method |
One of the character vector values of p.adjust.method. See help(p.adjust) for accepted values. |
Returns a list of data.frames.
Model Details |
A summary of the model details. |
ANOVA |
The second is the results of the ANOVA, including SS if they are requested in sQuote(detailed). |
Levene's test |
The third can be Levene's test for equality of variances if there are only between-subjects factors. |
Descriptives |
If they are requested, the last item is descriptive statistics: N, mean, SD. |
Tests of Simple Main Effects |
Output from the sme (Simple Main Effects) function. See help(sme) for details. |
Bill Altermatt
Bakeman, R. (2005). Recommended effect size statistics for repeated measures designs. Behavior Research Methods, 37 (3), 379-384.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.