Databases are rich with concealed information which can be used for intelligent decision making. Classification and association rule mining are vital to such practical applications. Thus, if these two techniques are somehow integrated would result in great savings and conveniences to the user. Such an integrated framework is called associative classification (AC). This integration is carried out by focusing on a specific subset of association rules whose consequent contains only class attribute. Several studies in data mining have shown that AC is superior to other traditional classification algorithms due to its numerous favorable characteristics such as readability, usability, training efficient and excellent accuracy. Hence, various AC techniques are studied with its pros and cons. However, AC suffers from a drawback that large number of rules is produced as an output. Now, utilizing all these rules for analysis would be computationally expensive. This paper studies various pruning and evaluation methods that are employed to produce qualitative rules. Further, the paper empirically evaluates associative classification technique considering various parameters.