APPENDIX 5
Memorandum from the Centre for Research
on the Wider Benefits of Learning
1. EXECUTIVE
SUMMARY
— The Centre for Research on the Wider
Benefits of Learning (CRWBL) is one of several DfES research centres.
CRWBL is a joint venture between the Institute of Education and
Birkbeck College with a specific remit to investigate the "non-economic"
benefits of learning.
— Through qualitative research, we
have identified that learning leads to more open-minded and tolerant
attitudes that may be thought to be associated with support for
environmentalism.
— Quantitative research on environmental
attitudes indicates that pro-environmentalism decreases between
age 33 and 42.
— Pro-environmentalism at 42 is positively
correlated with anti-racism and pro-welfarism. However, it is
negatively correlated with political efficacy and the work ethic.
— Adult education has a substantial
effect on reversing the decline in environmentalism between ages
33 and 42. One course of academic or vocational education is nearly
sufficient to reverse the tendency for pro-environmentalism to
decline.
2. INTRODUCTION
As part of our ongoing programme of work funded
by the department (DfES), the Centre for Research on the Wider
Benefits of Learning (CRWBL) has examined the impact of lifelong
learning on sustainability in its broadest sense. CRWBL is a joint
venture between the Institute of Education and Birkbeck College
with the specific remit of investigating the non-economic effects
of learning. Environmental sustainability has not been a specific
aim of our research to date, we have some findings on this issue
that we thought may merit the attention of the committee. These
concern the effects of lifelong learning on pro-environmental
attitudes.
3. QUALITATIVE
RESEARCH
Through qualitative research involving biographical
interviews with 140 learners (Schuller, Brassett-Grundy, Green,
Hammond and Preston, 2002) we have identified the influence of
lifelong learning on a variety of social attitudes. For many of
these learners, learning had led to an increase in open-mindedness
and tolerance, changes which are certainly associated with sustainable
communities, if not sustainability in an environmental sense.
In particular, adult learning was associated with increased civic
participation in terms of schemes such as community health promotion,
regeneration or mutual aid. Learning provided individuals with
the confidence (self-efficacy) which they needed to become involved
in these projects.
These qualitative findings have been complemented
by quantitative research which has identified specific effects
of adult education courses on environmental attitudes,
4. QUANTITATIVE
RESEARCH
Through the use of the nationally representative
NCDS (National Child Development Study) we have examined how attitudes
to the environment change over time and how they are effected
by participation in adult education. The NCDS is a representative
longitudinal study of around 10,000 individuals born in one week
(March 1958) who have been repeatedly surveyed concerning their
attitudes and behaviours.
4.1 Changes in Environmental Attitudes
In the fifth sweep of NCDS (when the respondents
were 33) and in the sixth sweep (when they were 42) questions
concerning agreement with three environmental attitudes were asked.
Respondents were asked whether they were strongly in agreement,
agreed with, were uncertain about, disagreed or strongly disagreed
with each statement.
In table 1, we show the degree to which individuals
(on average) in each sweep agree with the environmental attitudes
statements on a scale where one represents strong disagreement
with the statement and five strong agreement.
Table 1:
AGREEMENT / DISAGREEMENT WITH ENVIRONMENTAL
ATTITUDES
| Age 33 | Age 42
|
| E1 Problems in the environment are serious |
3.90 | 3.70 |
| E2 Preserving the environment is the most important political issue today
| 2.87 | 2.79 |
| E3 We should tackle the environment even if this means slower economic growth
| 2.03 | 1.98 |
(Table shows mean agreement with each statement. Number of respondents
(N) is greater than 10,000 for all cells)
As can be seen in table 1, on average there is agreement
that problems in the environment are serious (attitude e1). The
average (mean) score for agreement with the statement is 3.90
at age 33 and 3.70 at age 42. However, there is less agreement
that preserving the environment is the most important political
issue today (attitude e2) with an average score of 2.87 at 33
and 2.79 at age 42. Moreover, there is (on average) disagreement
with the statement that we should tackle the environment even
if this would result in lower economic growth (attitude e3) with
a score of 2.03 at age 33 and 1.98 at age 42.
This indicates that although individuals in this sample believe
that problems in the environment are serious, they are by no means
the most important political issue or one for which a goal of
economic growth should be sacrificed.
As can also be seen in table 1, between age 33 and 42 there
is a decline in the average agreement with these statements. For
example, in terms of attitude E1 (Problems in the environment
are serious) mean agreement has fallen from 3.90 at age 33 to
3.70 at age 42. This shift in the mean level of agreement between
age 33 and 42 is true of all three environmental attitudes. Further
research, not quoted here, indicates that this shift is better
explained by cultural change over the period then by any effect
of maturation for the individuals in the sample ie pro-environmentalism
has declined.
4.2 Relationship between Environmental and other attitudes
By combining environmental attitudes E1 to E3 we may form
a scale of environmental attitudes. The resulting scale pro-environmentalism
was relatively stable for the respondents between age 33 and 42
declining slightly from a mean of 3.66 (N=10,808, s.d.=.65) to
a mean of 3.64 (N=11,277, s.d.=.66)
As the NCDS also surveys other attitudes, we may investigate
the relationship between pro-environmentalism and these other
responses. In table 2 (below) we examine the relationship between
pro-environmentalism and six other attitudes at age 42 : anti-racism,
pro-welfarism (the degree to which individuals support "left
wing" policies), authoritarianism, family values, political
efficacy (the degree to which individuals believe in the democratic
system) and work ethic. The relationship is expressed as the correlation
coefficient; a statistic which indicates the degree of correspondence
between two variables. The correlation coefficient ranges from
-1 (perfect negative correlation) to +1 (perfect positive correlation)
[1]
Table 2:
CORRELATION BETWEEN PRO-ENVIRONMENTALISM AND OTHER ATTITUDES
AT AGE 4
| Attitude | Correlation
| p-value |
| Anti-racism | 0.141 | P<0.0001
|
| Pro-welfarism | 0.136 | P<0.0001
|
| Authoritarianism | 0.003 |
0.751 |
| Family values | 0.012 | 0.213
|
| Political efficacy | -0.021
| 0.028 |
| Work ethic | -0.041 | P<0.0001
|
| | (N>10,000 for all cells)
|
As can be seen in table 2, pro-environmentalism is positively
associated with anti-racism and being pro-welfarist in attitude.
This would seem to be historically consistent with environmentalism
being associated both with tolerance and traditional left wing
values. The association between pro-environmentalism and anti-racism
suggests sustainability both in the environmental sense and in
terms of creating sustainable, tolerant communities. There are
no positive associations with authoritarianism and family values.
On the other hand, pro-environmentalism is negatively associated
with both political efficacy and the work ethic although the association
is very light. This would suggest that pro-environmentalism is
associated with slightly less trust in the political system and
with lower adherence to values associated with employment. Again,
there is some consistency here between environmentalism and rejection
of conventional forms of political influence and work. We emphasise,
though, that the correlations are very slight.
4.3 Effects of Adult Learning on Environmental attitudes
Through the NCDS we know not only how individual attitudes
have changed between age 33 and age 42, but also whether they
have undertaken adult education and also the types of learning
they have undertaken. By controlling for other individual life
circumstances such as previous educational qualifications and
social class we may isolate the effect of adult education on pro-environmentalism
through a technique known as Ordinary Least Squares (OLS) regression[2].
Table 3 shows the effects of different types of adult education
course on pro-environmentalism in terms of increasing pro-environmentalism
controlling for other factors mentioned (see Annex 1). We have
divided courses into four broad course types: academic, vocational,
work-related and leisure. Taking the example of academic courses,
each academic course that the respondent takes increases pro-environmentalism
by 0.035 standard deviations. This effect is unlikely to have
occurred only for our specific sample (p=0.041). This means that
taking one academic adult education course between ages 33 and
42 is nearly sufficient to counter-act the tendency for pro-environmentalism
to decline between 33 and 42. (As the mean effect on pro-environmentalism
is to decline by -0.04 between 33 and 42, the effect of one academic
course of +0.035 standard deviations is nearly enough to compensate).
As the effects are linear in this model[3],
taking two academic courses would have an effect of +0.070.
Similar results are obtained for vocational courses where
each course increases pro-environmentalism by 0.037. Smaller but
important effects are obtained for leisure courses. The effect
size is still substantial because individuals may take a large
number of leisure courses.
Table 3:
EFFECTS OF ADULT EDUCATION ON PRO-ENVIRONMENTALISM
| Type of course | Academic
| Vocational | Work-related
| Leisure |
| Effect | 0.035 | 0.037
| 0.001 | 0.014 |
| p-value | 0.041 | 0.001
| 0.670 | 0.069 |
| (N= 9366) | |
| | |
5. CONCLUSIONS
The findings provided here are part of a continuing programme
of research. In particular, we will be examining the impact of
learning on sustainability and attitudes in more depth through
forthcoming qualitative and quantitative projects. However, we
consider these findings to be substantive in that adult education
appears to affect community and environmental sustainability.
Further research is needed in order to consider the mechanisms
by which these processes may occur. At present, we are utilizing
a framework of various capitals. For example, learning may increase
environmental sustainability by enhancing human capital (skills,
knowledge), social capital (networks and trust) and identity capital
(feelings of esteem and self-worth). Learning may also increase
individual capabilities in terms of problem solving. These mechanisms
need further testing both in the field (qualitatively) and through
further analysis of data. However, most important for policy makers
is the finding that policy variables can impact on environmentalism
and related attitudes.
February 2003
Annex 1: Results of Regression
Table 4:
OLS REGRESSION OF VARIABLES ON PRO-ENVIRONMENTALISM
N=9364, R-squared=0.2951, Adjusted R-squared=0.2934
| Variable | Coef.
| Std. Err. | T |
P>½t½ | [95% Conf.
| Interval] |
| - + | | |
| | |
|
| t61 | 0.03466 | 0.0169646
| 2.04 | 0.041 | 0.001406
| 0.067915 |
| t62 | 0.03696 | 0.0109443
| 3.38 | 0.001 | 0.015512
| 0.058418 |
| t63 | 0.001022 | 0.0023919
| 0.43 | 0.669 | -0.00367
| 0.005711 |
| t64 | 0.014458 | 0.0079008
| 1.83 | 0.067 | -0.00103
| 0.029945 |
| o6ldv | -0.84682 | 0.0136211
| -62.17 | 0 | -0.87352
| -0.82012 |
| n33ses5 | 0.024843 | 0.0494343
| 0.5 | 0.615 | -0.07206
| 0.121746 |
| n33ses3 | -0.0668 | 0.0486659
| -1.37 | 0.17 | -0.16219
| 0.028598 |
| n33ses4 | -0.03279 | 0.0497841
| -0.66 | 0.51 | -0.13038
| 0.064798 |
| n33ses2 | -0.03429 | 0.0487465
| -0.7 | 0.482 | -0.12985
| 0.061261 |
| n33ses1 | -0.09403 | 0.0633743
| -1.48 | 0.138 | -0.21825
| 0.0302 |
| n33ses—m | 0.001169 |
0.0573698 | 0.02 | 0.984
| -0.11129 | 0.113626 |
| n33aca—1 | 0.032467 |
0.0312673 | 1.04 | 0.299
| -0.02882 | 0.093758 |
| n33aca—2 | 0.053612 |
0.0284121 | 1.89 | 0.059
| -0.00208 | 0.109306 |
| n33aca—3 | 0.100067 |
0.0377126 | 2.65 | 0.008
| 0.026142 | 0.173992 |
| n33aca—4 | 0.098771 |
0.037136 | 2.66 | 0.008
| 0.025977 | 0.171566 |
| n33aca—5 | 0.088605 |
0.0727149 | 1.22 | 0.223
| -0.05393 | 0.231142 |
| n33aca—m | -0.09903 |
0.1455483 | -0.68 | 0.496
| -0.38433 | 0.186278 |
| n33voc—1 | 0.002506 |
0.027968 | 0.09 | 0.929
| -0.05232 | 0.05733 |
| n33voc—2 | 0.007869 |
0.0275507 | 0.29 | 0.775
| -0.04614 | 0.061875 |
| n33voc—3 | 0.031082 |
0.0298258 | 1.04 | 0.297
| -0.02738 | 0.089547 |
| n33voc—4 | -0.00243 |
0.026585 | -0.09 | 0.927
| -0.05455 | 0.049678 |
| n33voc—m | (dropped) |
| | |
| |
| Female | -0.0074 | 0.0189769
| -0.39 | 0.695 | -0.04464
| 0.029753 |
| —cons | 3.018673 | 0.0684374
| 44.11 | 0 | 2.884521
| 3.152825 |
Note : Table 4 (above) shows the results of an OLS regression
of various variables (dependent variables) on pro-environmentalism
(independent variable). The first column shows the variables entered
into the regression. Variables prefixed t represent adult education
(t61 : academic, t62 : vocational, t63 : work-related, t64 : leisure).
O6ldv is the level of pro-environmentalism at age 33. Variables
prefixed n33ses represent social class. Variables prefixed n33aca
and n33voc represent levels of academic and vocational qualifications
as defined by the DfES. Female is a self-explanatory variable;
whereas cons is the value of pro-environmentalism when all other
variables are zero.
BIBLIOGRAPHY
Schuller,T., Brassett-Grundy, A., Green, A., Hammond, C.
and Preston, J. (2002) Learning Continuity and Change in Adult
Life. (Wider Benefits of Learning Paper No.3, Institute of Education
/ Birkbeck College, London)
Feinstein, L., Hammond, C., Woods, L. & Preston, J. (2003).
"The effects of adult learning on health and social capital
and cohesion" WBL Research Report, forthcoming
1
Although the correlations between pro-environmentalism and other
attitudes are close to zero (suggesting no correlation), the large
sample size often means that there is a high likelihood that the
correlation is significantly above or below zero. For example,
in terms of the relationship between anti-racism and pro-environmentalism
the correlation is +0.141, suggesting a positive relationship.
The probability that we are mistaken in inferring that this relationship
is significantly different from zero is given by the p (probability)
value. Here, there is only a 1 in 10,000 chance (p=0.0001) that
we are mistaken in inferring that anti-racism has a positive correlation
with pro-environmentalism. Back
2
By standardizing our data prior to performing the regression we
have made the results easier to interpret. Through standardization
we find that the change in pro-environmentalism between ages 33
and 42 was ¸0.04. This means that, on average, individuals
moved 0.04 standard deviations1 away from the mean in terms of
their pro-environmentalism, becoming less pro-environmentalism
over time. This is a small change. One would expect roughly 60%
of individuals to move up to 0.3 standard deviations either way. Back
3
Linearity of this effect is assumed, but is considered elsewhere
in the ongoing work of the Centre (Feinstein, Hammond, Woods &
Preston, forthcoming, 2003). Back
| 
