How to analyse scale data?

#1
Hi all,
For my dissertation I have done a likret scale questionnaire, 1-5 (1=strongly disagree, 2=mostly disagree....) about body image (10 questions aimed at physical appearance, 10 aimed at what influences body image) I have separated it in to 2 age groups, young females and mature females, I want to see if there is a difference in opinion in the 2 age groups (eg are young females more influenced in their body image than mature females). I have put my data in to spss (all 100 completed questionnaires). in the variable view... Row 1 - age group = 1='young' 2='mature' and split this file(?). Then gone on to each question (Q1=1-strongly disagree, 2-mostly disagree ....Q2=......Q3=....).

I first proposed to do a 2 way independent ANOVA but the results made no sense in the data output, I have been searching all day to what I should do but I am now even more confused as I have the questions set to ordinal but then found people to say it should be nominal as it is a likert scale(?) Then people saying it is parametric so t-test should be done but then some say it is non-parametric so should be Mann-Whitney test or chi-square! I have tried all different tests but non are working so can anyone tell me what should be done?! I also tried computing variables for the physical appearance and the influence questions to still be even more confused :(

Also Should I group the 5 point scales to 3 (eg 1=agree, 2=uncertain, 3=disagree)?

Sorry this is long, I wanted to include as much information. If possible could any answers be step by step instructions? I understand to a certain extent then I get lost lol!
Thank you any help much appreciated :)