Hello dear 

I think you shall split age variable in spss software it located in data option in spss you can do this if you split age variable.

You said age was your dependent variable.  I seriously doubt that.  I think that age is an explanatory variable, and pain is the outcome (dependent) variable.  Do you have people's actual ages?  If so, I would start by inspecting a scatter-plot with X = Age and Y = Pain score.  If the relationship looks linear, fit a linear regression model.  Do not needlessly carve age into age groups, as this entails loss of power, and it imposes an artificial step function to the fitted values. I.e., everyone within the same age group has the same fitted value (the group mean), and two people just on opposite sides of an age cut-point can have very different fitted values despite having a tiny age difference.  See slides 24 & 25 in the presentation linked below, for example. 

HTH.

Presentation Free Advice for Power-Hungry Researchers: Do Not Categorize ...

The three groups of age are independent variable, and the scale of pain is the dependent variable.

you can do ANOVA test.

Since and respondent will have only one value in the pain scale. Then using One way ANOVA you can compare mean pain difference among the three age groups. this will tell you only if there is mean difference or not. However, if you are interested with age group are significant with which age group you can conduct the post analysis such as Tukey pairwise test.

Thank you

Dear Jonathan,

given the way you describe your design, it looks like a One-way ANOVA should be able to do the job (3 factors of age = independent variable, i.e. the variable you manipulated) and felt pain level as your dependent variable. This will look for an overall effect (are the three groups different from each other?) Post-hoc tests will then tell which individual groups perform different from each other.