Sibthorpe, J., Paisley, K., Gookin, J., and Furman, N. (2008) ‘The Pedagogic Value of Student Autonomy in Adventure Education’, Journal of Experiential Education, 31 (2): 136-151.

Student autonomy is widely believed to be educationally important ‘the student belief that he or she has some sense of meaningful control’ (137).  Student autonomy seems to lead to both academic and personal outcomes including ‘perceived competence, self esteem, creativity, and conceptual understanding’ (137).  Supporting student autonomy means providing more choice, a clear rationale, and ‘perspective taking’.  Teachers need to provide students with actual occasions where they can choose, get them to discuss the reasons why they might want to learn something, and offer a chance to show that they can understand and sympathise with students positions and perspectives.  Specifically, teachers might make sure they are ‘asking for opinions, listening to answers, allowing others to talk, encouraging effort, and praising progress (Reeve, 2005)’ (137).

Activities need to be meaningful, something that makes an actual difference.  This can vary between individuals, but there might be common agreement on matters such as not spending a great deal of time deciding what to have for lunch, as opposed to ‘being able to plan the content, timing, flow and execution of a full day of programming’ (138). Student voice can be encouraged ‘by allowing participants to initiate, design and participate in activities of their own choice and...  having access to adults in the programme who listen to and respect youth’s opinions’ (138).

The idea of a ‘challenge by choice’ is often incorporated, allowing participants to choose and regulate their own levels of involvement.  Such empowerment is claimed to lead to ‘self efficacy…  Life effectiveness…  Leadership, communication skills and outdoor skills’.  Actual practices might include requiring participants to be leader of the day, set goals and lead expeditions [sounds like what I did as a 14 year old naval cadet] (139).

However student led expeditions and independent travel (‘autonomous student expeditions ‘ or ASE) can bring significant risks, and withdrawing supervision has been challenged.  This paper sets out to see if autonomy does lead to the claimed development outcomes, especially empowerment, and to attempt to assess physical risks. [It is very ambiguous with the former -- see below]  Participants were on a national outdoor leadership course.  Data were collected using an instrument designed for the course and based on course objectives, measuring ‘perceived gains in leadership… [and].. outdoor skills… [using a ] retrospective pretest/post test format’ (141).  Personal empowerment was measured by a specially designed ‘Characteristics of the Experience Scale (Sibthorpe, 2001)’ [must look this up]. 

Risk was estimated by examining ‘the injury and evacuation statistics’ on a particularly particularly well recorded course There was also ‘a two item lie scale’, and data were also collected from the instructors.   Chi square scores were calculated to compare actual numbers of injuries and evacuations with those expected.  None of the actual numbers of injuries etc were statistically different than what one might expect from chance alone, but ‘inspection’ seems to show that there are no more incidents during unaccompanied parts of the course (142).  Overall, ‘the contention that the risks from these approaches are not acceptable was not supported by the data in this study’ (144). 

The substantial results are summarised on page 142, and are quite difficult to follow.  Pretest scores were compared with post test, and the data analysed [in some way I don’t understand—hierarchical linear modelling?].What seems to be happening is that models are constructed so that particular factors of interest can be used to predict responses on tests elsewhere.  For example one model tested the predictive power of a personal empowerment score on subsequent scores .  Another one took the number of days of ASE as the predictor for subsequent scores.  For technical reasons other models were constructed which included no predictor variables in order to explain the variance in the ‘two outcome variables: leadership and outdoor skills’ (142). So here, the outcome variables seem to be scores, possibly self rated, on leadership and outdoor skills, measured before and after the course took place.  Other technical refinements are difficult to follow.  For example if the course variable (number of days of ASE) explained a lot of the variance, additional terms were added to the models—‘the predictors and cross-level interaction terms’ (142): I must say I don’t understand this.  A residual variance was also calculated, allowing ‘for the variability of the regression slopes’.  Variables that did not explain a significant portion of the variance were removed from the model.  Apparently, both the null models, or the other models mentioned above, found a significant amount of variance attributable to course differences—in particular that ‘between 11 and 15% of the variance in the outcomes could be attributed to course differences’ [so does that word ‘significant’ mean statistically significant or theoretically significant?  Whatever it means, a maximum of 15% of the variance seems pretty small beer to me!].

Then a personal empowerment score was added as an additional variable.  While they were there, the team also did some ‘cross-level interaction’ to see if personal empowerment and days of ASE were themselves interacting (142).  You will be relieved to know that the interaction between these variables ‘was not significant’.  However, both variables on their own ‘were significant predictors’ of gains in both leadership and in outdoor skills, in both cases and for both variables at a .01 level of significance.  Confusingly, despite the remarks about low levels of interaction above, ‘days of ASE can predict empowerment.  Participants who have experienced more days …  of ASE on their courses reported higher levels of empowerment’ (143).  The authors think that ‘the empowering mechanism of ASE’ is the active ingredient here [but more confusion follows—ASE also has empowerment, so it is hardly a surprise that it predicts empowerment measured on another scale?].

The course trained outdoor leaders with some experience,  and ‘program participants may be markedly different than other adventure outdoor programme participants’.  In addition, some other measures are ‘self reported perceptions and are, therefore, subject to participants’ biases’ (145).  Ideally additional factors associated with the course should have been controlled as potential mediating variables as well [not specified --selection for the course? duration?].  Nevertheless, ‘the authors believe that ASE are pedagogically valuable’ (145) [no doubt before they even did the research].

The specific teacher behaviours above ‘are all immediately translatable as leader behaviours in adventure education’ (146).  However, leaders still have a responsibility to ‘align the menu of choices with program goals’ (146).  One useful technique appears to be having instructors think out loud, discussing a particular decision with other instructors in the presence of participants.  Perspective-taking seems particularly important as well, especially when instructors exercise skills habitually, and may not remember how they had learned them—‘how can a complex task be broken down into manageable steps’ (146).  Allowing students to set goals can be used, for example where they choose to focus on a particular skill ‘such as fishing or climbing’, and the instructors can allow more time for these activities.  Transferring responsibility enables students to take the perspective of instructors and explore of rationales, especially the  ‘leader–of–the–day’ technique (147).  While these experiential techniques are not new, ASE does offer ‘a rather unique opportunity for providing student autonomy’ (147).

However, ‘the absence of instructors should not be automatically associated with autonomy support’ (147), and without their interventions, groups left to their own devices would probably not develop ‘equitable opportunities for autonomy’ (147).  Instructors may also have to build their own teaching skills and judgment.  One technique that might minimise risk is ‘remaining close to but unobserved by groups throughout the day and camping relatively nearby at nights, and intentionally meeting the student group at hazards’ (147).


Selected refs

Sibthorp, J. (2001). Development and validation of the characteristics of experience scale for use in adventure based programming [Abstract]. Proceedings of the 1999 Symposium on Leisure Research (p. 60). Ashburn, Virginia: National Recreation and Park Association.
Sibthorp, J. (2003). An empirical look at Walsh and Golins’ adventure education process model: Relationships between antecedent factors, perceptions of characteristics of an adventure education experience, and changes in selfefficacy.
Journal of Leisure Research, 35(1), 80–106.

Sibthorp, J. & Arthur-Banning, S. (2004). Developing life effectiveness through adventure education: The roles of participant expectations, perceptions of empowerment, and learning relevance. Journal of Experiential Education 27(1), 32–50.

back to host page