Means Test

[Marzieh Goodarzi]

Hi everyone
I know that perhaps I should not ask this question here.
But I hope you can help me.
I want to do the "Means Test" using the people's records data, and my time is very limited. What should I do to get started?
If you know the right source to learn this in brief, please introduce.
thanks a lot.

[Carlo Lazzaro]

See -mean-.

[Marzieh Goodarzi]

thanks. but i mean : A means test is a determination of whether an individual or family is eligible for government assistance, based upon whether the individual or family possesses the means to do without that help.

[Carlo Lazzaro]

Marzieh:
thanks for clarification.
You may want to search for community-contributed commands by Stephen Jenkins

[Leonardo Guizzetti]

I could be mistaken, but I think means tests are not statistical tests, but rather a question of whether a person's income is above or below a specific threshold. You would need to know (at least) what these various thresholds are. For example, low-income thresholds often vary by location and number of people within a household and/or number of dependents.

[Carlo Lazzaro]

Yes, it was my thought, too.
Possibly, Stephen Jenkins has contributed with some user-written programmes.
That said, googling with -jarman deprivation index Stata- gives back some potentially interesting entries.

[Marzieh Goodarzi]

Leonardo :
Yes, that's what I mean.
thanks.

[Marzieh Goodarzi]

Carlo Lazzaro:
Thank you for your guide

[Stephen Jenkins]

I support Leonardo Guizzetto in #5. There won't be a generic program available because means tests (a.k.a. income tests) are a set of eligibility rules that vary by country and time period (and family type). You need a "tax-benefit calculator", either writing your own code to encapsulate the rules that apply in the particular context, or using a tax-benefit microsimulation model if there is one available (and there often isn't!)

[Marzieh Goodarzi]

Stephen Jenkins:
thanks for your guidance.

In fact, I have access to bank account information for two consecutive years, and I want to divide people into three groups with a certain income level.
But I must minimize the systematic error or ... in identifying people based on their income.
Now, my question is, how should I minimize these errors?
And by which method can I test the estimates?
(Other information I have access to them is about having a home and having a car.)