These activities
are what our findings revolve around. (3) Our third step was to
review each transcribed and typed interview data sentence by
sentence to extract quotations associated with the key themes
representing the compensating mechanisms that the third party
provided. (4) We used triangulation to check the validity of our
data. We triangulated data from interviews and observations; the
data reported in our results represent data from both sources. Our
research strategy allowed us to demonstrate the validity of the
data in three ways. First, we developed initial conceptual cate-
gories and some preliminary explanations from the data. Second,
we added new categories as new information emerged and
dropped some conceptual categories. We accomplished this by
comparing new information we uncovered from the data to prior
explanations we derived from that data. This process of iteration
continued until we believed that we were not getting any new
information (Corbin and Strauss, 2008). Finally, we collected data
from all the stakeholders: the bridge and its leadership, corpora-
tions and supplier firms.