Outdoor education/Program length

From Wilderdom

Jump to: navigation, search


Program length


A common program model/design question is around what is the optimal length for an outdoor education program.

For some earlier thoughts on this topic, see Factors Which Influence the Effects of Outdoor Education Programs

This page provides some additional notes about quantitative research evidence examining the relationship between program length and participant outcomes.

The overall finding is that program length tends to be a weakly positive predictor of measured outcomes. Thus, longer programs are probably more effective, but this is a relatively weak effect.

There are several problems/limitations with the available research evidence, including that:

  1. There is a common belief amongst outdoor education practitioners that "longer programs are better"
  2. Length is often confounded with other factors (e.g., program type) which limits the usefulness of available data.
  3. There is no qualitative data on program length
  4. One, practical, underused method to help determine optimal program length is conduct a program evaluation survey at the end of the program which includes a question asking participants to indicate whether they would have liked the program to be shorter, about the same, or longer.

Basically, longer programs are probably more effective, but this effect appears to be relatively weak.

Of course, these are very general findings, so may or may not apply to a specific program/context or population group.

Nevertheless, my sense is that it may be more about focusing on the program structure, components, processes etc. for program design/improvement rather than program length per se.

Research findings

References can be found via PhD.

Program length has been found by several studies to be weakly, positively associated with larger participant outcomes (e.g., Cason & Gillis, 1994; Hattie et al., 1997; Sibthorp, Paisley, & Gookin, 2007). However, program length is often confounded with program type, making previous research imprecise. Duration effects may not be linear (e.g., outcomes might follow a law of diminishing returns). Program duration may be a proxy for psychological growth processes which are slightly more likely to manifest in a longer program. However, whether the psychological growth processes occur is highly likely to be influenced by other participant, group, activity, and program philosophy characteristics. Thus, a particularly well-conducted program may be able to effect notable change in a relatively short duration. Program length, then, is expected to positively predict outcomes, but only to a small extent.

More detailed findings from specific studies are described below.

Cason (1993; Cason & Gillis, 1994) conducted a meta-analysis of the effects of adventure-based programs, based on data representing 43 studies, 147 effects, and 2,291 adolescents. They found that program length had a significant, weak, positive, linear relation with outcome (r = .17; durations ranged between 36 and 5400 hours; Mdn = 54 hours (3 weeks)).

Hattie et al. (1997). Hattie et al. (1997) conducted a meta-analysis of the effects of adventure education programs, with data representing 96 studies (1,728 effects and 12,057 participants) conducted between 1968 and 1994. Hattie et al. (1997) found that 36% of the variance in the outcomes could be explained by three significant independent variables: (a) the organisation running the program (highest results were for Outward Bound Australia programs ); (b) program length (programs greater than 20 days were the most effective); and (c) age of participants (adults had larger effect sizes than adolescents).

Russell and Sibthorp (2004) conducted a multi-level study of wilderness therapy programs for youth at risk. Dependent variable data were based on the pre-post change total scores for adolescent emotional and behavioral symptoms as measured by 523 adolescent self-reports and 372 parent reports using the Youth Outcome Questionnaire. The group/program-level explained 14% of the variance and 86% of the variance was at the individual level. Program length explained 67% of the group/program-level variance, and 9% of the total variance, with longer programs being more effective. For parent response data, the group/program-level explained 20% of the variance (with 80% of the variance at the individual level). Program length explained 46% of the variance in the sample data, and 9% of the total variance. In general, there was little in this study to indicate that program length ranging between 3 and 26 days were a useful predictor of life effectiveness outcomes. Longer programs were weakly, but non-significantly, associated with more positive short- and long-term outcomes. Importantly, the analyses in the current study accounted for the potentially confounding effects of Program Type and Participant Age which may have affected estimates of the effect of Program Length in other studies (e.g., Cason & Gillis, 1994). Previous meta-analytic research had suggested a weak, positive relationship between outdoor education program length and participant outcomes. The current study found weak, non-significant trends in this direction for each program type (with the effect of participant age partialled out). The lack of significant relation between length and outcome could be due to this study exercising more control over potentially confounding variables (program type and participant age). The findings could also be influenced by range restriction (3 to 26 days).

Sibthorp, Paisely, and Gookin (2007) analysed National Outdoor Leadership Schools (2004; N = 596) participants in 66 groups. Outcomes were retrospectively perceived changes for three psychosocial areas (communication, leadership, and small group behavior) and three outdoor areas (judgment in the outdoors, outdoor skills, and environmental awareness). Length of program was a positive, significant predictor for five out of six outcomes. Paisley, Sibthorp, Furman, Schumann, and Gookin (2008) recently conducted similar analyses with 2005 and 2006 data (N = 1,228) which largely replicated and confirmed the 2004 findings.

Neill (2008) examined the relationship between program length and life effectiveness outcomes for 3,640 Outward Bound Australia program participants during the 1990s. Program length had small, positive, partial correlations with short-term changes in overall life effectiveness (with the effect of Participant Age partialled out). The overall correlation between length and life effectiveness outcome was .06, explaining less than 1% of the variance. Program length parameter estimates were mostly positive, but non-significant, for each of the eight life effectiveness factors examined. Program length was most important for Active Initiative (0.16), Self Confidence (0.13) and Time Management (0.12), with only the Active Initiative effect being statistically significant (t > 1.96, p < .05). Program Length did not significantly predict short- or long-term outcomes, with the exception of positively predicting short-term Active Initiative outcomes, a small effect. Program Length coefficients were very small, positive, and non-significant for short-term outcomes. Overall, program length was a small, positive predictor, of marginal significance. Thus, there was only weak, largely non-significant evidence to support the conjecture that longer programs are more effective; the vast majority of the variance in outcomes appears to be independent of Program Length. These results are smaller than the correlation of .17 between Program Length (Cason and Gillis, 1994) but somewhat similar to findings by Hattie et al. (1997). The lack of notable relationship between program length and life effectiveness outcomes in the current study could indicate that program length is a less salient feature of outdoor education programs than is commonly thought. More dynamic process-type variables could be more accurate predictors. For example, individual development may be more dependent on, or related to, group processes (Ewert, 1992; Ewert & Heywood, 1991; McAvoy, Mitten, Stringer, Steckart, & Sproles, 1996) than to the duration of the intervention per se. Whilst a minimal length of time is probably necessary, all outdoor education programs generally seek to provide a “complete” experiential cycle, for example by following “hero’s journey”-type stages (Bacon, 1987; Campbell, 1968; Lord, 1987; Loynes, 1999, 2003). Environmental contrast (perceived degree of separation or contrast with home environment; Fabrizio & Neill, 2005), program intensity, and individual coping strategies could also be more salient, alternative constructs.

Bowen and Neill (2013) conducted a meta-analysis of 197 studies of adventure therapy participant outcomes (2,908 effect sizes, 206 unique samples). Program length was not a significant moderator of outcomes. A break-down of effect sizes by program are shown in the table below:

Program Length NSamples NES ES
1-2 days171220.44
3-7 days271620.53
8-14 days241610.41
15-21 days272800.35
22-45 days484180.42
46-80 days334580.39
81-150 days14500.62
151+ days11990.71

What is a program?

Walsh and Golins (1976) suggest in their theoretical conceptualisation of Outward Bound programs, that there is much more going on in a program than its length:

"Outward Bound ... is not something which is necessarily 23 days in length; has training, expedition phases, a solo, final expedition, and a marathon; or is conducted by an Outward Bound school ... Nor is it the values, such as: self-preservation, self actualization, perseverance, initiative, reflection, experimentiveness, etc. ... Instead we are talking about the structures, components, and conditions whose presence and interaction insure that an experience is conducive along the lines of Outward Bound (i.e., aids the persons involved in reaching the objectives intended.). (Walsh & Golins, pp. i-ii)"

This seems to suggest an explanation for why simply extending program (without also improving the dynamic structure, components, conditions, and processes) will not necessarily improve outcomes. It may be more effective, for example, to improve recruitment and screening processes, enhance instructor training, and/or review and adjust program activities, than simply to extend a program.

Personal tools