My study consists of a conceptual framework with seven constructs measured in seven point Likert scale. According to pilot study, only three constructs have Cronbach alpha value more than 0.9 while others remain between 0.7 and 0.8.
Cronbach alpha value represents the reliability of the scale. 0.9 value is good hoping that you have followed the procedure of scale development. Validity would be more important for research.
Please also see the size of your sample and if it is representative of the population.
Such a high Cronbach's alpha indicates high internal consistency, i.e. a concept is being reliably assessed. This does not indicate the scale is unidimensional - you need to run factor analysis to assess that. Also, high Cronbach's alpha does not indicate validity. When developing a psychometric instrument you also need to consider validity (face and construct) plus test retest reliability.