Cronbach Alpha Testing for tau equivalence This question is related to a previous question I opened
I have a scale I wish to apply Cronbach alpha to. One of the assumptions of Cronbach's is the tau equivalence assumption which according to my review of the literature is barely ever achieved. I was initially using the package tau.test in the  coefficientalpha package but it was crashing when I was using a construct with two questions (see the original question). 
The previous recommendation was that I should use the psych package but after going through the documentation I cant seem to find a test which will tell me if all questions contribute equally to the construct I am testing
Can anyone point me in the right direction of how I can make this test?
In essence I would just want a flag that is basically 1, all questions contribute equally to the construct or 0, the questions don't equally contribute. Long term, I want to show my reasoning why I used omega to measure scale reliability instead of the more ubiquitous alpha
If this question is more for StackOverflow please do move it
Thank you for your time
 A: I don't think that you can test the modelfit using the psych package (or I don't know how), thus you have to use the lavaan package.
I use the data that you mention in your other question for an example:
library( lavaan )
library( semPlot ) # for graphics

# Data
x  <- round(rnorm(n=709 ,mean = 6, sd = 1))
x1 <-round(rnorm(n=709 ,mean = 5.6, sd = 1.3))


mydf <- data.frame(A1 = x,
               A2 = x1)

Tau-Congoneric Model:
You have to specify the model you want to test, using the variable name of your data:
model_con <- 'A =~ A1 + A2'

Than you will test you model using a confirmatory factor Analysis:
cfa_con <- lavaan::cfa(model_con, data = mydf, estimator = "ml")

A graphic can make it more clear:
semPlot::semPaths(cfa_con , whatLabels="stand") 


To see which model has a better fit you need to check the measures of fit which are stored in the cfa object. Then you can compare it with other models and see which fits the best:
lavaan::inspect(cfa_con,"fit")[ c(2,3,4,8,19,23) ]

Tau-equivalence
model_taueq <- 'A =~ 1*A1 + 1*A2
A1~a*1
A2~a*1'

cfa_taueq <- lavaan::cfa(model_taueq, data = mydf, estimator = "ml")

semPlot::semPaths(cfa_taueq, whatLabels = "stand") # the example data is worse for lavaan, thus no graphic can be made
lavaan::inspect(cfa_taueq,"fit")[ c(2,3,4,8,19,23) ]

