How much does a successful collaboration between a university researcher and a funding source affect the researcher's future preference of that source? To examine this relationship, another crucial variable is added, namely, the individual's level of dependency on funding sources. It is hypothesized that given a low level of dependency, if researchers have had a satisfactory experience with a funding source, they will be more likely to prefer collaboration with that source than if they did not have a satisfactory relationship. Given a high level of dependency, past experience will not make a difference on preference. The effects of past experience on people's preferences are examined based on a survey of American university researchers. The data show that past experience has a strong impact on one's preference, but this effect is true for all researchers, regardless of their level of dependency on external support. Implications for funding sources, and social scientists using resource dependency as the theoretical framework are discussed.