Shaosong Ou
Construct: E-Integration
 
 
 

Items:

For vertical integration, the following two measures are generally accepted:

Maddigan (1981):

Vertical Industry Connection (VIC).

Scherer and Ross (1990):

Value Added to Sales (VC) ratio.

 

For electronic integration, the following items are found:

Zaheer and Venkatraman (1994):

Electronic Integration is operationalized as the percentage of business (commercial premiums) directed to the interfaced carrier through the proprietary electronic channel. Specifically, the item used in the paper is "What's the percentage of business (commercial premiums) directed to the interfaced carrier through the proprietary electronic channel?", measured by percentage point number varying from 0 to 100. As can be seen, this operationalization does not directly measure what the original definition proposes (i.e., a vertical quasi-integration achieved through the deployment of dedicated computers and communication systems between relevant actors in the adjacent stages of value-chain), instead it measures what the operationalization definition proposes, another definition form.

Christiaanse and Venkatraman (2002):

In consistent with the above work, this study operationalizes electronic integration as "the percentage of the independent travel agency's business directed to the focal airline (American Airlines) whose information systems is dominant in the agency business operations". The data on market share does not follow a standard normal distribution, it was rescaled using a natural log transformation.

Other similar operationalizations include:

John and Weitz (1988):

"Forward integration" is operationalized as percentage of direct sales to end users.

Masten et al. (1989):

Backward integration is operationalized as the percentage of company's component needs produced under the governance of the firm.

Caves and Bradburd (1988):

A continuous measure of vertical integration is based on the share of industry output.

 

Not many validity measures have been found for this one-item construct. Zaheer and Venkatraman (1994) test its convergent validity by examining its strength of association with two other indicators - the number of carriers accounting for 80% of each agency's business as well as the number of carriers that each account for at least 2% of each agency's business. The correlation of EI measure with the former is -0.5847, and the correlation of EI measure with the latter is -0.4788, both significant at 0.01 level.