Page 201 - Šolsko polje, XXVIII, 2017, no. 5-6: Znanje, motivacija in pogoji učenja v luči mednarodnih primerjav TIMSS in PISA, ur. Barbara Japelj Pavešić in Klaudija Šterman Ivančič
P. 201
kozina, m. štraus ■ relationship between academic achievement and wellbeing ...
With included social dimension predictors, 12 percent of variance
in achievement in literacy domains was explained for reading and mathe-
matics and 14 percent for science. As in the previous model, the strongest
predictor is socio-economic and cultural status in all three domains. The
values of the regression coefficient for this predictor are also similar to the
values in the previous model.
Across all three domains, connectedness to the neighbourhood is
a significant predictor of achievement in literacy domains. This predic-
tor is negative; students who reported higher levels of connectedness to
the neighbourhood on average achieved lower scores on the three litera-
cy assessments than their counterparts. For reading literacy the regression
coefficient of -8.25 score-points indicates that students reporting a one
standard deviation higher scale value of connectedness to the neighbour-
hood on average scored 8.25 score points lower on the reading literacy as-
sessment than their counterparts with lower scale value of connectedness
to the neighbourhood but similar values of other predictors. Standardized
regression coefficient (beta) indicates that this represents approximately 9
percent of the overall standard deviation in the reading literacy scores in
the student population. For mathematical and scientific literacies, the re-
gression coefficients of this predictor were similar; students reporting a
one standard deviation higher scale values of connectedness to the neigh-
bourhood on average scored 8.47 points lower in mathematics and 9.35
points lower in science than their counterparts reporting lower scale val-
ues of connectedness to the neighbourhood but having values of other
predictors similar. In both, mathematics and science literacies, the differ-
ence in scores represents approximately 10 percent of overall standard de-
viation of the respective scores in the population. One would expect that
students that feel more connected to their important others have higher
achievement in literacy domains as well, but this was not the case in our
data. As we can see from the data, low achieving students are more related
to important figures outside school (in the neighbourhood). One hypoth-
esis would be that due to the low literacy achievement they feel less con-
nected to their school and school peers and more to their neighbourhood
peers. This is supported also with our results showing that having more in-
timate friendships is related to higher literacy in reading and feeling more
connected to teachers in the school is a predictor of higher achievement in
math. Students reporting more frequent intimate friendships score on av-
erage, 6.10 points higher in reading than their counterparts with less inti-
mate friendships. This accounts for approximately 7 percent of the overall
standard deviation of reading scores in the population. Similarly, connect-
edness to teachers predicted mathematics literacy with the change of 4.57
199
With included social dimension predictors, 12 percent of variance
in achievement in literacy domains was explained for reading and mathe-
matics and 14 percent for science. As in the previous model, the strongest
predictor is socio-economic and cultural status in all three domains. The
values of the regression coefficient for this predictor are also similar to the
values in the previous model.
Across all three domains, connectedness to the neighbourhood is
a significant predictor of achievement in literacy domains. This predic-
tor is negative; students who reported higher levels of connectedness to
the neighbourhood on average achieved lower scores on the three litera-
cy assessments than their counterparts. For reading literacy the regression
coefficient of -8.25 score-points indicates that students reporting a one
standard deviation higher scale value of connectedness to the neighbour-
hood on average scored 8.25 score points lower on the reading literacy as-
sessment than their counterparts with lower scale value of connectedness
to the neighbourhood but similar values of other predictors. Standardized
regression coefficient (beta) indicates that this represents approximately 9
percent of the overall standard deviation in the reading literacy scores in
the student population. For mathematical and scientific literacies, the re-
gression coefficients of this predictor were similar; students reporting a
one standard deviation higher scale values of connectedness to the neigh-
bourhood on average scored 8.47 points lower in mathematics and 9.35
points lower in science than their counterparts reporting lower scale val-
ues of connectedness to the neighbourhood but having values of other
predictors similar. In both, mathematics and science literacies, the differ-
ence in scores represents approximately 10 percent of overall standard de-
viation of the respective scores in the population. One would expect that
students that feel more connected to their important others have higher
achievement in literacy domains as well, but this was not the case in our
data. As we can see from the data, low achieving students are more related
to important figures outside school (in the neighbourhood). One hypoth-
esis would be that due to the low literacy achievement they feel less con-
nected to their school and school peers and more to their neighbourhood
peers. This is supported also with our results showing that having more in-
timate friendships is related to higher literacy in reading and feeling more
connected to teachers in the school is a predictor of higher achievement in
math. Students reporting more frequent intimate friendships score on av-
erage, 6.10 points higher in reading than their counterparts with less inti-
mate friendships. This accounts for approximately 7 percent of the overall
standard deviation of reading scores in the population. Similarly, connect-
edness to teachers predicted mathematics literacy with the change of 4.57
199
