**3.2 Instrument validity and instrument reliability**

There are several types of validity measures such as face validity and construct validity. Campbell and Fiske [66] propose two types of validity: convergent and discriminating validity. Convergent validity is measured by average variance extracted for each construct during the reliability analysis that should be 0.5 (50%) or better. **Table 2** shows the reliability values for the various constructs with variance extracted, and all the values are above 50%, thus providing a sufficient evidence of convergent validity. Similarly, Cronbach's α [67] for the constructs ranging from 0.80 to 0.90 further indicate a sufficient level of reliability. In general results show that both validity and reliability requirements are met.

## **3.3 The sample**

A questionnaire was sent to 129 SMEs according to a random sampling plan. The SMEs were selected from a key business directory of Brunei (www.goldpages.com). Out of these 127 organizations, 70 organizations responded, and responses from 67 organizations were retained as they were filled by the top management; three were dropped because of the fact that it was not filled as per instructions. This makes the response rate of 52% sufficient for the survey of SMEs especially in a small market of Brunei Darussalam.
