47
if the
author made recommendations, and if the findings were otherwise disseminated (table
6). The level of application of findings also was a marginally significant factor, with findings
applicable at a single species level having higher rates of implementation
than at a multiple
species level.
Table 6. Minimum adequate model for factors influencing the implementation of research findings (logistic
regression (df=423). Only the factor levels of significance are shown in detail P(>|z|), otherwise only the
significance of the factor as a whole is represented P(>|Chi|)
Factor
Factor level
Direction
P(>|z|)
P(>|Chi|)
Dissemination
Motivation (research)
Author capacity
Recommendations
Research ongoing
Funding local NGO
Species level
Status of research country
Year
Author capacity:status
Dissemination:Year
Yes
Scientific
Academic
Yes
Yes
No
Yes
Developed
-
-
-
Positive
Negative
Negative
Positive
Positive
Negative
Positive
Positive
-
-
-
0.32
3.58e-06
0.63
0.003
6.72e-06
0.008
0.009
0.02
-
-
-
3.572e-07
2.219e-09
0.02
1.453e-03
4.417e-06
1.871e-03
0.01
0.21
0.46
0.04
0.03
Although the year of publication had no significant impact on
the uptake of findings in
univariate analysis (table 5) or indeed as a main effect (table 6),
there was a significant
interaction on the uptake of findings combined with dissemination. In 2000, whether or not the
results were disseminated had no impact on uptake of findings
whereas in all other years,
including 2005, there was a largely significant interaction between
dissemination and uptake
of findings (fig. 20). This could suggest a time lag in implementation
of research findings
directly from the research paper, but is more likely a function of smaller sample size in 2000,
as levels of implementation in 2001 were statistically different from 2000, but not 2005.
