This study evaluates the relationships between states of the four channel flow model and adventure seeking traits among whitewater kayakers using a modified Experience Sampling Method. Four study hypotheses were concerned with determining whether the interaction between adventure seeking and the four channel flow model predicts differences in dimensions of subjective experience (flow indicators): (H^sub 1^) Among flow indicators, a three-dimensional structure to include Affect and Activation, Cognitive Control, and Intrinsic Freedom will be confirmed; (H^sub 2^) Adventure seeking will have a significant interaction with the flow channel when predicting a dimension of Affect and Activation; (H^sub 3^) Adventure seeking will have a significant interaction with the anxiety channel when predicting a dimension of Cognitive Control; and, (H^sub 4^) Adventure seeking will have a significant interaction with flow and apathy channels when predicting a dimension of Intrinsic Freedom.

Questionnaires were administered on-site to 52 whitewater kayakers on the Cheat River in West Virginia at eight sites varying in river difficulty (Class I-V). Data were analyzed at the level of experience (n = 409 experience observations) rather than per respondent. Statistical analyses (principal axis factoring and hierarchical linear modeling or HLM) confirmed a three-dimensional structure of flow indicators, and that the interactions of adventure seeking and the channels of the flow model were significant predictors of an Intrinsic Freedom dimension. Although the adventure-seeking trait was a significant predictor of the Affect and Activation dimension, this dimension and the Cognitive Control dimension were not significantly predicted by interactions with channels of the flow model.

The study results suggest several conclusions that fully support two of four hypotheses. The three-dimensional structure of flow indicators was validated by factor loadings, and latent measures were well supported by reliability coefficients. Jackson and Marsh (1996) and others have generally confirmed a high number of dimensions rather than combining indicators to form only several dimensions. The application of a ninedimensional structure would be incompatible or impractical to use with many analytical (i.e., HLM, Lisrel) and methodological strategies (i.e., experience sampling 27 flow indicators on a river]. Furthermore, the confirmation of the three-dimensional structure implied that the use of a higher number of dimensions (i.e., nine) of subjective experience may be redundant in representing subdimensions of the overall structure.

Although results did not fully support the second hypothesis, when controlling for the four channel model, adventure-seeking significantly contributed to the prediction outcome of the Affect and Activation dimension. The finding that kayakers had a positive affect and were more mentally excited while in states of flow and anxiety as opposed to states of boredom is consistent with prior conceptualizations and empirical findings of Csikszentmihalyi (1982), Csikszentmihalyi & Csikszentmihalyi (1988), and others. Furthermore, this interpretation is theoretically consistent with Zuckerman's (1994) notion of Thrill and Adventure seeking, which proposes that individuals who are high in adventure seeking seek out those experiences that arouse and excite the self. It should be noted, however, that the small to medium effect sizes for these pairwise comparisons suggest a need for improvements in explanatory power.

With a moderate level of explanatory power, results also demonstrated that, when controlling for the four channel model, the interaction of positive levels of adventure seeking coupled with entering the flow state generated significantly more intrinsic freedom than during a state of apathy. This finding is consistent with the theoretical and empirical expectations of Csikszentmihalyi (1982), Csikszentmihalyi & Csikszentmihalyi (1988), and Jackson and Marsh (1996), and suggested that when interacting with adventure-seeking, optimal experiences were indicated by the freedom to choose involvement in kayaking the Cheat Canyon and the desire to repeat this intrinsically motivating experience. Likewise, this finding is consistent with Moneta's (2004) HLM results, which indicated that higher intrinsic motivation is associated with the flow state.

Two interpretations can be made for the finding that the Cognitive Control dimension had no significant fixed effects. One is that this dimension is less characteristic of optimal experience in adventure settings than previously theorized (Csikszentmihalyi, 1982). The second interpretation is that Cognitive Control is not being empirically tapped because there continue to be methodological and operational issues within the four channel model that contribute to measurement error.

The findings that adventure-seeking is either interactive with channels of the flow model or able to predict flow indicators while controlling for the model, suggests several implications for researchers and practitioners concerned with the link between the flow state and personality traits. Researchers should attempt to further unveil the salience of the interaction between adventure seeking and the flow model in predicting intrinsic freedom. To accomplish this task, it is recommended that investigators employ research designs that follow the experience from the river to environments with increased exposure to extrinsic activities and where experiences are at their least adventurous points in time and recover once again. Critical to practitioners, this design would allow for a realization of the long-term benefits and outcomes of optimal experience. Agencies (i.e., Outward Bound, NOLS) should also understand that these findings stress the psychological importance of providing elements of adventure in wildland settings by demonstrating that optimal experience in wildland environments is an interactive process with client personalities. Thus, future researchers examining experiential education settings should attempt to determine whether facilitating experiences in the adventure, without simultaneously facilitating for client personality, limits levels of perceived intrinsic freedom necessary to achieve the most optimal of experiences.