Agriculture Reference
In-Depth Information
Table 4.4 Factors associated with practising UA (logistic regression model)
Beta
Standard
error
P
value
Exponential
beta
Suburb type: high density ¼ 1, low
density ¼ 0
3.03
1.05
0.004
20.71
Gender of head of household: male¼ 1,
female¼ 0
1.50
1.38
0.28
0.22
Age of head of household (continuous)
0.02
0.02
0.38
0.98
Educational level of head: 1¼primary,
0 ¼more than primary
2.07
1.03
0.04
7.92
House ownership: 1 ¼owner;
0 ¼non-owner
1.51
0.51
0.003
0.22
Household size (continuous)
0.28
0.14
0.05
1.33
Formally employed members: 1¼any
formally employed, 0¼none
0.88
0.88
0.32
2.42
Informally employed members: any
informally employed¼1, 0 ¼none
0.49
0.91
0.59
1.64
Constant
18.88
49.64
0.70
15.91
Overall percentage correct 89.9%. Significant results are shown in bold.
In this study, households headed by males were, on average, significantly
larger than households headed by females.
Though female-headed households were smaller in size than male-headed
households, they had significantly higher dependency ratios. Maxwell et al
(2000) noted a similar finding in Kampala, Uganda, in the early 1990s.
Households headed by females contained a much larger proportion of relatives
and grandchildren than those headed by males. Female heads tend to be older
(mean age 50.9 years) than their male counterparts (mean age 45.9 years).
Results for the logistical regression on UA are shown in Table 4.4. A
correct prediction value of 89.9 per cent was obtained. This meant that the
variables used were very good at statistically predicting the observed outcome
of practising UA. Suburb type (P
¼
0
:
004), educational level of head
(P
¼
0
:
04), household size (P
¼
0
:
05) and house ownership (P
¼
0
:
003) were
significantly associated with participation in UA (see also Table 4.4).
Suburb type
Households in the high-density suburbs were more likely to practise UA than
their counterparts from low-density suburbs, a finding consistent with
ZIMVAC (2003), Mwakiwa (2004) and Mudimu et al (2005). From the odds
ratio given by exp
, those from high-density suburbs were 21 times more
likely to practise UA than their low-density counterparts.
