How do I recode this independent, categorical variable

  • Context: Undergrad 
  • Thread starter Thread starter cachemony
  • Start date Start date
  • Tags Tags
    Independent Variable
Click For Summary

Discussion Overview

The discussion revolves around the recoding of a categorical independent variable (INST) in SPSS for the purpose of calculating bivariate correlations with a continuous dependent variable related to measures of democracy. Participants explore how to transform the original variable into two separate variables representing Presidential and Parliamentary systems while excluding mixed democracies and dictatorships.

Discussion Character

  • Technical explanation
  • Debate/contested
  • Homework-related

Main Points Raised

  • Noah seeks to recode the INST variable into two distinct variables for analysis, specifically for Presidential (PRES) and Parliamentary (PARL) systems, while excluding mixed democracies and dictatorships.
  • One participant suggests creating a single categorical variable (PRES) that equals 1 for Presidential systems and 0 for Parliamentary systems, while excluding observations where INST equals 0 or 2.
  • Noah clarifies the intention to analyze correlations separately for Presidential and Parliamentary systems.
  • Another participant points out that excluding INST values of 0 and 2 would lead to perfect correlation between the two new variables, as they would be inversely related.
  • There is a suggestion that if Noah does not exclude the 0's and 2's, it would be possible to define both PRES and PARL as independent variables.

Areas of Agreement / Disagreement

Participants express differing views on whether to exclude mixed democracies and dictatorships from the analysis, leading to a lack of consensus on the best approach to recoding the variable.

Contextual Notes

The discussion highlights the implications of variable exclusion on correlation analysis and the potential for perfect correlation between the newly defined variables if certain values are excluded.

cachemony
Messages
2
Reaction score
0
I’m student at the University of Arizona. In my political science class, I need to use SPSS to calculate bivariate correlation between a categorical independent variable (Presidential or Parliamentary system) and a continuous dependent variable (various measures of democracy)

I’m using the Przeworski data set. My original independent variable in this set is called INST, which designates each state as a dictatorship 0, parliament democracy 1, mixed democracy 2, and a presidential democracy 3.

My question is: How do I recode this independent, categorical variable (INST) into two different variables (Presidential and Parliamentary), so it excludes the mixed democracies and dictatorships and allow me to run a correlation with my continuous dependent variables? So, I want to be able to recode INST into Presidential and Parliament variables and run them separately to see how each one correlates with my continuous dependent variables?



Thanks,
Noah
 
Physics news on Phys.org
If I understand this correctly, you do not want to code INST into 2 different variables. You want a single categorical variable, say, PRES which is = 1 if INST = 3, and PRES = 0 if INST = 1. Before doing this, however, you should exclude all observations where INST = 0 or 2.
 
I want to recode INST so I can run correlations of Presidential and Parliment systems separate from each other against my dependent variables?

Thanks,
Noah
 
If you exclude INST = 0 or 2, you cannot define two independent variables, e.g. PRES and PARL, because they will be perfectly correlated with each other: PRES + PARL = 1 or PARL = 1 - PRES.

OTOH, perhaps you are not aiming to exclude the 0's and the 2's; in which case you can define two independent variables PRES and PARL.
 
Check out www.youtube.com/mycsula for our newly completed series of online SPSS (now IBM Statistics) video tutorials.

Learn how to: define variables, enter data, run a paired samples t test, perform frequency analysis, use scripting and syntax files, create and edit charts and graphs, and much more.

Training modules cover a wide range of topics aimed at beginners to more advanced users. All videos include examples and instructions to help you with whatever area of research or analysis you are working on.

You can find these videos at www.youtube.com/mycsula. Check them out and subscribe to our channel today!

Thanks,

The ITS Online Training team

www.youtube.com/mycsula
 

Similar threads

Undergrad Handling categorical variables in R
  • · Replies 2 ·
Replies
2
Views
2K
Undergrad Count and Categorical Variables...
  • · Replies 3 ·
Replies
3
Views
2K
Undergrad How to show these random variables are independent?
  • · Replies 1 ·
Replies
1
Views
3K
Undergrad Linear regression and random variables
  • · Replies 30 ·
2
Replies
30
Views
5K
Graduate Do outliers exist in categorical data and how can they be detected?
  • · Replies 6 ·
Replies
6
Views
5K
Undergrad Regression on extracted factors
  • · Replies 2 ·
Replies
2
Views
1K
Graduate How to find which variables w, x, y.... 'most' influence z?
  • · Replies 7 ·
Replies
7
Views
2K
Graduate Independent variable that is a distribution?
  • · Replies 1 ·
Replies
1
Views
2K
Graduate How Many Hidden Variables Would It Take to Model Photon Polarization?
  • · Replies 45 ·
2
Replies
45
Views
7K
Undergrad Thermodynamic functions and state variables
  • · Replies 8 ·
Replies
8
Views
2K