anova.hopit | R Documentation |
Perform the likelihood ratio test(s) for two or more hopit
objects.
## S3 method for class 'hopit' anova(object, ..., method = c("sequential", "with.most.complex", 'with.least.complex'), direction = c("decreasing", "increasing"))
object |
an object containing the results returned by a |
... |
an additional object(s) of the same type. |
method |
the method of ordered model comparisons. Choose |
direction |
determine if the complexity of listed models is
|
a vector or a matrix with the results of the test(s).
Maciej J. Danko
print.lrt.hopit
,
lrt.hopit
, hopit
.
# DATA data(healthsurvey) # the order of response levels decreases from the best health to # the worst health; hence the hopit() parameter decreasing.levels # is set to TRUE levels(healthsurvey$health) # Example 1 --------------------- # fitting two nested models model1 <- hopit(latent.formula = health ~ hypertension + high_cholesterol + heart_attack_or_stroke + poor_mobility + very_poor_grip + depression + respiratory_problems + IADL_problems + obese + diabetes + other_diseases, thresh.formula = ~ sex + ageclass + country, decreasing.levels = TRUE, control = list(trace = FALSE), data = healthsurvey) # a model with an interaction between hypertension and high_cholesterol model2 <- hopit(latent.formula = health ~ hypertension * high_cholesterol + heart_attack_or_stroke + poor_mobility + very_poor_grip + depression + respiratory_problems + IADL_problems + obese + diabetes + other_diseases, thresh.formula = ~ sex + ageclass + country, decreasing.levels = TRUE, control = list(trace = FALSE), data = healthsurvey) # a likelihood ratio test lrt1 <- anova(model1, model2) lrt1 # print results in a shorter form print(lrt1, short = TRUE) # or equivalently lrt.hopit(model2, model1) # Example 2 --------------------- # fitting additional nested models model3 <- hopit(latent.formula = health ~ hypertension * high_cholesterol + heart_attack_or_stroke + poor_mobility + very_poor_grip + depression + respiratory_problems + IADL_problems + obese * diabetes + other_diseases, thresh.formula = ~ sex + ageclass + country, decreasing.levels = TRUE, control = list(trace = FALSE), data = healthsurvey) model4 <- hopit(latent.formula = health ~ hypertension * high_cholesterol + heart_attack_or_stroke + poor_mobility + very_poor_grip + depression + respiratory_problems + IADL_problems + obese * diabetes + other_diseases, thresh.formula = ~ sex * ageclass + country, decreasing.levels = TRUE, control = list(trace = FALSE), data = healthsurvey) # sequential likelihood ratio tests # model complexity increases so direction = "increasing" anova(model1, model2, model3, model4, direction = "increasing", method = "sequential") # likelihood ratio tests of the most complex model with the rest of the models anova(model1, model2, model3, model4, direction = "increasing", method = "with.most.complex") # likelihood ratio tests of the least complex model with the rest of the models anova(model1, model2, model3, model4, direction = "increasing", method = "with.least.complex")
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.