1. ## Which statistical test?

Hi

I am currently writing my bachelorthesis. The subject is bankruptcy models. There are model 1, 2 en 3 that can predict a bankruptcy of a company. I would like to investigate if the accuracy ratio will effected by the period. The accuracy ratio is calculated as follows: number of good prediction/total number of prediction.

For example:

0 = not bankrupt
1 = bankrupt
with ( ) = actual
with out ( ) = prediction
sample size = 21

The companies in period 1 en 2 are same

Periode 1 (sample 1)
Model 1: 0(0), 0(0), 0(0), 0(0), 1(0)..... = in this case 80% accuracy
Model 2: 1(0), 1(0), 1(0), 0(0), 0(0)..... = in this case 40% accuracy
Model 3: 0(0), 0(0), 0(0), 0(0), 0(0)..... = in this case 100% accuracy

Periode 2 (sample 2)
Model 1: 0(0), 0(0), 0(0), 0(0), 0(0)..... = in this case 100% accuracy
Model 2: 1(0), 0(0), 0(0), 0(0), 0(0)..... = in this case 80% accuracy
Model 3: 0(0), 0(0), 1(0), 1(0), 1(0)..... = in this case 40% accuracy

I would like to indicate if the differencies (in accuracy ratio) in period 1 and 2 are Statistical significance

My question: Which statistical test should I use to test the statistical significance in period 1 and 2.

2. ## Re: Which statistical test?

Assuming you have counts or frequencies, you have two categorical traits - So a simple chi2 test would test the significance of this.
P1model1 vs P2model1
P1model2 vs P2model2 etc..

Disclaimer (im not a statistician, a biologist)

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts