Questions about Second-Level Analysis

Hello, 

I am new to the CONN and hope to get some answers about second-level analysis. 

Our project is to investigate the effect of sleep quality and HIV status on resting-state connectivity. 

Here is how I construct the general linear model: 
1. group contrast: HIV [1 1 1 1 .....0 0 0 0] and control [0 0 0 0 .... 1 1 1 1]
2. sleep quality: continuous variable, adjusted mean (orthogonalized)
3. interaction between HIV and sleep quality: HIV [1 1 .. 0 0] * sleep, e.g., [4.3 3.7 0.2 ..... 0 0 0 0 0]
4. covariates: sex (male [1] and female [0], e.g., [1 0 0 1 1 0 0 0 1...]), age (continuous variable)

I selected all variables for second-level analysis. Is the model correct? Should I use [1 1 .. 0 0] or [1 1 ..-1 -1] for the interaction term? What should I select for the main effect of HIV status, Difference HIV > Control [1 -1 0 0 0 0] or effect of HIV [1 0 0 0 0 0]? 

When I examined the ROI-to-ROI connectivity for effect of interaction, I noticed that there were blue lines between two regions and red dots on the targeted region. The beta value is positive. What does it mean? If blue line means negative correlation, why is the beta value positive? I have attached the graph here. Could you please explain the interpretation of ROI-to-ROI connectivity in terms of the blue/red lines and blue/red dots? Thank you so much!