DjalmaPessoa

this is above my head ...

It is not clear to me the step: variances

are you using the formula in: https://stats.stackexchange.com/questions/78479/how-to-run-chi-squared-test-on-imputed-data with the correction? On Mon, May 29, 2017 at 3:55 PM, Guilherme Jacob wrote: > I think it's equation 2.2 in page 74....

Where is defined the distribution of the test statistic in 2.1. What are the number of degrees of freedom in the numerator and in the denominator of the F distribution?

you mentioned that: The test was meant for the cases when m = 3 ? In your example k= (c-1)x(r-1)= 16 , the value of k has no influence in...

It is good you believe in your books..., Multiple imputation is bit misterious and magic to me... For k=16 if the missingness fraction is high, m=3 sounds magic! The test...

I just said I'd rather use Wald's test!

Example using Lumley's suggestion: library(mitools) data.dir

To replicate the estimates in https://biblioteca.ibge.gov.br/visualizacao/livros/liv101390.pdf we have to use the microdata in ftp://ftp.ibge.gov.br/Trabalho_e_Rendimento/Pesquisa_Nacional_por_Amostra_de_Domicilios_continua/Anual/Microdados/Dados/PNADC_2016_entr1_20171201.zip

The estimates mentioned by Miguel also used the microdata in ftp://ftp.ibge.gov.br/Trabalho_e_Rendimento/Pesquisa_Nacional_por_Amostra_de_Domicilios_continua/Anual/Microdados/Dados/PNADC_2016_entr1_20171201.zip This stresses the need to download and read them using lodown.