My draft model has statistics that show not "good fit" in that RMSEA = 0.104 (higher than 0.07), CFI = 0.758 (lower than 0.90)and SRMR = 0.105 (higher than 0.05). Am adviced to improve it using modification indices.
The data will not be impacted. I will try to improve it, for example, by taking only variables with factor loading >=0.5, I had used 0.4 cut-off. I will post revised statistics thereafter to see the change.
After applying modification indices, the RMSEA = 0.029 (target is 0.90)and SRMR = 0.087 (higher than 0.05). In addition, the co-efficient between Indepedent latent variable and the Dependent latent variable is 0.1481 - is this too low an association?