# SPSS - coding confusion ( Likert scale)

#### alex8888

##### New Member
Hi)
I face a problem with coding Likert scale .
I read that if u have, let's say 4 statements and Likert scale( strongly disagree 1 to strongly agree 5) and one of the statment is negative u have to do reverse coding.
But what if all my statements are negative what should i do???
Example:
What is your attitude to programme after the deal?(likert scale 1-5)
1.I start to trust programme less
2.It looks less favorable now
3. I stop recommend it to friends

How should I code variable 1: that '1' is strongly disagree( like normal variant) or '1'strongly agree???

I really get confused=(

#### Injektilo

##### New Member
If all of your statements are negative, you don't need to do anything.

The recoding is only done to ensure that all statements go in the same "direction". Otherwise if you're using FA/PCA, correlations, regressions, ANOVA or whatever, you may end up getting confused by the presence of negative signs. But if all of your statements are already in the same "direction", than you can go ahead with what you have.

Note that if you're looking to run a regression on some other variable as the dependent variable that is taking a positive direction (that is, opposite of that of your statements), you will end up with negative beta scores. You may then want to recode all those statements' code values to avoid that from happening.

#### alex8888

##### New Member
Many thanks for answer !!! I get it!
But about regression, if I get negative beta it means that X and y are negatively correlated ? So it should be no problem or better to avoid negative beta?

#### Injektilo

##### New Member
Yes, it means that X and Y have a negative relationship, but you just don't want to confuse yourself if it's due to X having a positive spin and Y having a negative spin. In that case, the relationship is actually positive, and the negative value is reflective of the 2 statements taking opposite directions. It's easy to confuse yourself in those situations.