Regression with different scales variables (need help)

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

haidrich
Posts: 3
Joined: Sun Jan 12, 2014 7:15 pm

Regression with different scales variables (need help)

Postby haidrich » Sun Jan 12, 2014 7:37 pm

Hello,

I'm working here on my research proposal for my alma mattera. I'm in accountancy science field, however, I need to deal with some statistics on this occassion. Would be very grateful if you could help me.

I have an idea to built this kind of regression model:

Y= x1 + x2 + x3

The research question is:
Do audit prices, non audit services provided for audit client and auditor's tenure has an impact on audit quality?

Observation - financial reporting of particular company. Let's say, we have 40 companies and their audited financial reporting.

Y (proxy for audit quality) - count of errors in financial reporting. I guess it would suitable to group the errors found in each observation. For instance, range from 1 to 5 (where 1 - from 0 to 2 errors, 2 - from 2 to 4... and so on.). Or it could be left in relative scale.

x1 - price of audit services of financial reporting. It could be left in model in a relative scale, or coded as 1 - price above the average price in the sample, 0 - price bellow the average price in the sample

x2 - non-audit services provided for the audit client. 1 - there were other services provided, 0 - there where no other services provided.

x3 - the tenure of auditors. Like: 1 - the same auditors are auditing the financial reporting 3 or more years in the row, 0 - the same auditors are not auditing the financial reporting more or equal 3 years.

That's my assumptions how to encode the data. Could you express your opinion how the data should be encoded actually, what kind of regression should I use here and what statistical tests woudl be suitable for evaluating the results?

Thank you very much
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Regression with different scales variables (need help)

Postby GerineL » Mon Jan 13, 2014 9:17 am

It is better to keep data that is actually on a scale as scale data, as you throw away information if you aggregate the data. Also, it is not nessecary for your question to aggregate the data.

if you would just keep them in scale, you can use multiple regression.
haidrich
Posts: 3
Joined: Sun Jan 12, 2014 7:15 pm

Re: Regression with different scales variables (need help)

Postby haidrich » Mon Jan 13, 2014 10:27 am

Hi,

Thanks for your answer.

To be sure that I understood you well, you suggest to keep all variables in ratio scale, like in these examples:

Observation 1: A 2200 USD (price) + B 1610 USD (value of provided non audit services) + C 2 years of tenure = 8 errors
Observation 2: A 1500 USD (price) + B 0,00 USD (value of provided non audit services) + C 6 years of tenure = 2 errors
......

?

thanks
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Regression with different scales variables (need help)

Postby GerineL » Mon Jan 13, 2014 10:30 am

yes that is my suggestion.

Please note that the type of answer you get is not in terms of "if you have more than 3 years experience", but in terms of whether or not (and how strongly) the years of experience matter.


In addition, if you do have the data entered as scale, you always have the option to later recode it into other variables and do other (additional) analyses. However, if you enter it as 0 - 1 - 2 etc, you don't have the option to convert the data back to scale.
haidrich
Posts: 3
Joined: Sun Jan 12, 2014 7:15 pm

Re: Regression with different scales variables (need help)

Postby haidrich » Mon Jan 13, 2014 10:37 am

Okay, understood you, thank you.

However, I believe there is a risk that many observations among the sample will have values 0,00 in the variable ,,Non audit services provided". Will it decrease the reliability of the analysis?
GerineL
Moderator
Posts: 1477
Joined: Tue Jun 10, 2008 4:50 pm

Re: Regression with different scales variables (need help)

Postby GerineL » Mon Jan 13, 2014 2:24 pm

if there is not enough variance that could be a problem, but this can be tested.

again, you are safe to enter your data in scale, and in about 1 minute you can transform this to any other variable you want for your analysis (e.g., make it 0-3 etc).

You have most possibilities left if you enter your data as scale.

also, if you are worried because they are all 0: you can always standardize or center your vars to get rid of that problem.

Who is online

Users browsing this forum: No registered users and 2 guests

cron