Dummy moderator variable impact on relationship IV and DV

[lanems]

Hi

I am wondering someone on this forum can answer this question to relation to following

We have a research model where DV_1 is perceived usefulness we have a number of independent variables
IV_1
IV_2

Moderator variable is gender (dummy variable Female = 1, Male = 0) MV_1 - Directly MV_1 with DV_1 has a very low beta coefficient 0.039 (ns)

However when MV_1 is included as moderator with:

IV_1 * MV_1 = beta coefficient of interaction effect is 0.13
IV_2 * MV_1 = beta coefficient of interaction effect is -0.171

Q1. which Interaction effect term generation should we use for a dummy variable and a continuous variable in moderator interaction effect?

1. Use original indicator values for multiplication
2. Standardise indicator values before multiplication
3. mean-centre indicator values before multiplication (My reading of literature on use of moderator variables in regression analysis suggests this option)

Q2 If the direct effect of the moderator variable such gender is not significant but the interaction effect of the moderator variable and independent variable is significant is this worth reporting?

Q3 Is it better to do multigroup analysis using the dummy variables to create a number sub sets of the whole data set and compare groups female versus male in our model ?

Q4 We have 8 variables 22 indicators and N = 208 and we have a categorical variable four business groups but one of business groups has only on 15 responses in the data set

Can we still run a multigroup analysis on that business group?
Or are the numbers too low and may be we can run an ANOVA to see if there are differences across business groups for DV_1 perceived usefulness

Any advice on above much appreciated

Kind regards Michael