MANOVA, when independent variable is not multi-level and P value significance varies!

My research has 4 independent variables and 4 dependent variables. My hypotheses are constructed in such a way that 1 independent variable's effect size is checked on 4 dependent variables.

My independent variables are not continuous but they are measured on a 7 item likert scale. And my dependent variables are also measured on a 7 item likert scale. Independent variables are 4 subscales of collective self esteem scale. Public, membership, private and identity. Dependent variables are 4 shopping dimensions *brand consciousness, recreational, impulsive and fashion consciousness

MANOVA is an automatically good choice for this. My independent variables are not built around categories and groups e.g. gender and neither it is multi-level like income level. Hence post hoc tests are not relevant in this case.

My concerns are as follows,

1. whether MANOVA'S without post hoc tests is a good choice and as effective? Which values should Ii specifically look upto. Wiki's lamba, partial eta square and between subjects?

2. Secondly is there other alternative statistical test for the above mentioned situation.

3. Thirdly, a grave concern is, when I ran regression analysis and ANOVA of one independent variable on one dependent variable, some P values were insignificant, but in MANOVA those P values were significant. What does it signify, is there a concern here?

Ii would be thankful if above mentioned queries are answered Sincere regards