Creating overall score, aggregating by manager ID.

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

Creating overall score, aggregating by manager ID.

Postby » Fri Jan 09, 2009 2:08 am

Hi, I hope someone can help with this.

I am working with a raw data file, attempting to get the data to combine in the same way a system combines it.

There are 29 items and I need an overall score for each manager. Typically I would just aggregate manager ID and Compute a Mean(item 1, item 2...etc) to get the overalll score. (or compute an overall score for each individual in the raw data file (using the same compute) and then aggregating by mgr id). Unfortunatley, neither of these are the way the system creates the overall score.

I believe the system combines the scores like this.

Lets say a manager has 2 employees, and they fill out a survey with 5 questions about their mgr, likert scale type items. This manager is then associated with 10 data points. The system computes the mean by adding up all 10 data points and dividing them by 10 to come up with the overall score.

I'm trying to figure out a way to get SPSS to do this. Currently, I can compute a variable so that i can add up items 1-5 for employee 1 and 2 separately, and if I aggregate by manager id, then I'm essentially averaging the 2 employee average responses, which will give me a slightly different average in some cases.

Also, I can aggregate my manager id first, getting an average of all the items, then compute the means of item's 1-5, which also gives me a slightly different mean.

Is there any way to do the type of mean I need in the raw data file?

I've tried spliting files by manager ID, but that doesn't change anything when I compute a variable....

any ideas?
Posts: 391
Joined: Sun Dec 28, 2008 5:55 pm
Location: Belo Horizonte, Brasil

Postby Fierce » Fri Jan 09, 2009 3:24 am

If I got it right, your data is somethin like this:

Code: Select all

MNGR  Q1   Q2   Q3   Q4   Q5
1     x1   x2   x3   x4   x5
1     y1   y2   y3   y4   y5
2     i1   i2   i3   i4   i5
2     j1   j2   j3   j4   j5
And you want SPSS to calculate (x1+x2+x3+x4+x5+y1+y2+y3+y4+y5)/10 (same for i,j), right?

Compute a new variable, say nvar1 = Q1+Q2+Q3+Q4+Q5 (no means, just the sum) and then aggregate but choose SUM and not MEAN for the aggregating function. After that, compute a new variable, say meanvar = nvar1_sum/10.


Code: Select all

[(x1+x2+x3+x4+x5)/5 + (y1+y2+y3+y4+y5)/5]/2 = [(x1+y1)/2 + (x2+y2)/2 + ... + (x5+y5)/2]/5 = (x1+x2...+y4+y5)/10 = THE MEAN
In words: I if got it right, everything you did is the same. If the values are different, should be a decimal-places problem.

Who is online

Users browsing this forum: No registered users and 1 guest