Socioeconomic status, academic achievement and timeuse in eight-year-old Portuguese children Vítor Teixeira Orlanda Cruz University of Porto Faculty of Psychology and Educational Sciences Abstract Time-use literature considers that the way children spend their time influences their cognitive and social development. On the other hand, socioeconomic status has been found consistently to be a powerful variable explaining academic achievement. This study examines whether the relation between SES and academic achievement could be mediated by children’s time-use. Participants were 113 children, aged eight and nine (53 girls and 60 boys) living and studying in the Metropolitan Area of Porto. Children’s time-use was collected through a time-diary, adapted from the Child Development Supplement of Panel Study of Income Dynamics (University of Michigan). Only number and type of activities will be considered in this study. Children’s academic achievement was assessed using the academic competence factor of the Portuguese adaptation of the Social Skills Rating System (SSRS). Preliminary results show that the number of daily activities, time spent in social activities and outdoor unstructured activities and reading time are related both with SES and academic achievement.

1. Introduction Throughout the years and despite of the Portuguese educational system successive reforms, we’ve have been assisting to the maintenance of school failure. The research about this phenomenon shows that the socioeconomic status (SES) is systematically and strongly associated to children’s academic achievement. In fact, “student’s schooling outcomes, particularly their academic outcomes, have long been related to their SES” (Ma, 2000, p. 337), and many studies go on searching the relation between these two variables (Ma, 2000; Marjoribanks, 2001; Walker, Petrill & Plomin, 2005; Caldas & Bankston, 1997). Considering the SES as a multidimensional concept, there are several ways to define it (Crane, 1996; Hoff, Laursen & Tardif, 2001; Walker et al, 2005; Okpala, Okpala, & Smith, 2001; Ma, 2000). Among those, Hoff and collaborators detach the parent education. For this authors this variable is “stable, and, of all the socio-economic variables, maternal education (ME) has been found to be most strongly associated with parenting” (Hoff et al., p. 234). So, in the research presented here we are interested in studying this specific association between the ME and the children’s school performance. It seems to be a hard to break cycle, a sort of social inheritance that the mothers leave to their children – those who have mothers with higher levels of education, are more likely to present school success and to achieve themselves higher education levels. In the same line of Walker and collaborators (2005), we try to understand the way specific aspects of the environment mediate children’s school performance, aiming to identify environmental risk factors that can be modified in order to improve the school achievement. It seems relevant to understand how different levels of ME influence the children’s school performance, that is, what processes are involved in this relation. This kind of knowledge would allow us to design intervention plans directed, not to the increase of the families’ school level, but to provide children environment conditions necessary to increase of their school achievement. In this study we put the hypothesis that the effect of the ME in the school performance is mediated by variations observed in children’s time use. In the social sciences there is a tradition in the measurement and analysis of the way adults use their time; for example, economists view time as a unit of “human capital” (Larson & Verna, 1999). In our study we consider children’s time as a resource that can be used “for the development of a much wider range of faculties besides economic ones, including social competencies and dispositions related to healthy emotional adjustment” (Larson & Verna, 1999, p. 702). In fact, is our conviction that the complexity and the contents of the activities can be seen as an indicator of the psychological development of children (Bronfrenbrenner, 1979). Everyday life, his activities and routines are full of interest from the psychological and the developmental point of view (Palacios, Hidalgo & Moreno, 2001), defining the opportunities to learn and practice specific competences (Bronfrenbrenner, 1979, Larson & Verna, 1999, McHale, Couter & Tucker, 2001). We believe therefore that the way children use their time could be an important aim of intervention if we intend to promote their developmental level and, particularly, their school performance. The main goal of this study is to test a mediation effect of children’s time use in the association between the ME and the academic achievement. We did not find in the literature studies that analyse this specific mediation effect, so we will present a brief review about the relation between relevant variables: ME, academic achievement and the several dimensions of time use. We want to understand which variations of children’s time use, associated with the ME, are also related to the children’s academic achievement.

1.1 Maternal Education and Academic Achievement Parents influence children’s progress both in and out of school (Gilliam, Gerla & Wright, 2004). For these authors, “the more children grow up in a literate environment, where books and talk are a part of their daily lives and where their parents interact with them trough oracy and literacy experiences, the better their chances for success in school” (p.226). In this line, the authors refer a report (Texas Literacy Council, 1989, in Gilliam et al, 2004) that relates the school failure in schools of the state of Texas with the huge rate of illiteracy and the low levels of schooling of the parents of school-age children. For Wright, Diener and Kay (2000), in impoverished places the infants “grow up in families who don’t understand the importance of talking and reading to preschool children. Their parents think that learning starts when they enter school. Because they lack these critical early childhood experiences, the chances of being successful in school are remote and the cycle of poverty continues” (p.100). On the other side, Zady and Portes (2001) point that decades of research about the relation between family characteristics and children reading difficulties suggest that the children whose parents had reading difficulties, are themselves at a higher risk for the same difficulties. Also in what concerns the children's performance in mathematics it is possible to note the effect of ME. Being ME one of the measures of the SES referred by White (1982), the author referred several studies that demonstrate that the ME has an independent effect in children's marks in mathematics. Finally, in what concerns the resources provided to children, Walker and collaborators (2005) refer that parents with higher levels of education are more apt and aware of the importance of putting their children in schools with better educational resources, showing that the schooling of the parents influences “a vast set of variables, like the school performance, the school resources, the school' s and professors quality, the group of peers, the size of the class, and the amount of time in school." (p. 68). 1.2 Maternal Education and Time Use There is strong empirical evidence that the parents’ education is one of the main factors to contribute for qualitative variations in children's time use (McHale et al, 2001). These authors point that “children’s activities were associated with social class (SES) indices, with parents’ education being the most consistent correlate” (p.1769). Higher levels of education are related to more accelerated rhythms of life and to a more intense participation in cultural and social life, which leads to a more active use of the free time (Carvalho, 2001). In families with a higher level of education the children spend more time reading, or hearing someone that reads to them (Hofferth & Sandberg, 2001), perhaps because the parents read more for themselves, there are more books available at home and the children are more encouraged to read. For example, mothers with more years of schooling read twice the time for their children than the mothers with lower levels of education (cf. Carvalho, 2001). Also in the families with higher levels of education children spend longer periods of time studying (Hofferth & Sandberg, 2001, Lopes & Coelho, 2002). Sheldon (2002) notice that high income families, as well as higher education families, are more likely to be

involved in the education of their children. These parents seem to be more preoccupied with their children's academic achievement and, because of that, they spend more time supervising children’s learning and homework tasks (Yeung, Sandberg, Davis-Kean & Hofferth, 2001; Cosden, Morrison, Gutierrez & Brown, 2004), being more aware of the influence of parental involvement in children’s development. On the other hand, Carvalho (2001) establishes that although women with higher levels of education lead busier professional lives having less time available to spend with their children, they also tend to organize better their free time, having a more active social life. McHale and collaborators (2001) refer several studies with school-age children demonstrating that children from lower SES families spend more time in informal and unstructured activities; watch more television; read less; and are less involved in sports activities. They also refer that lower SES children “spend more time playing outdoors (riding bikes, playing in playground equipment) and less time in quiet indoor play with toys and games” (Newson & Newson, 1976; in McHale et al, 2001, p. 1765). On the other hand, in higher income families and in families with higher levels of education, children watch less television (about half an hour during a week) (Hofferth & Sandberg, 2001; Lopes & Coelho 2002; Mc Hale et al, 2001) and are more collaborative in household activities (Carvalho, 2001). 1.3 Time Use and Academic Achievement There are several dimensions of children's time use that, in a positive or negative way, seem to be associated with children's academic achievement. As a next step, we will proceed to a brief review of the activities carried out by children in their daily routines that appear to be related to school performance. Time spent playing outdoors and hanging around is associated with poorer school grades (McHale et al., 2001), this association being probably related to the fact that children do so “with peers in settings unsupervised by adults” (McHale et al., 2001, p. 1774). In the same line, Parke and Buriel (1998) refer several studies indicating that parents of delinquent and anti-social children do less supervision and monitoring of the children’s activities, than parents of non-delinquent children. These data allow us to conclude that less parental supervision and monitoring associated with more time spent by children playing outdoors and hanging around, is related to children’s less-adaptive functioning, which includes both poorer school grades and more conduct problems (McHale et al., 2001). Time spent watching television has been related to lower results in tests of cognitive achievement, and also to less time spent in activities such as reading and studying (Hofferth & Sandberg, 2001). Stressing this idea, Shin (2004) advances three possible explanations for the negative association between watching television and school achievement: more time watching television implies less time to carry out other more stimulating activities; watching television requires less mental effort reinforcing a more passive mental functioning; watching television encourages impulsive behaviours and causes attention problems and consequently drives to the decrease in the school performance. Time spent reading is positively associated with children's academic achievement (McHale et al, 2001; Hofferth & Sandberg, 2001) and, in earlier ages, predicts academic results in more advanced school levels (McHale et al, 2002; Cappella & Weinstein, 2001). Home computer use is also positively associated with the school performance (Subrahmanyam, Grrenfield, Kraut & Gross, 2001), although the authors reinforce the need of “more studies to better understand those effects" (p. 17).

Also time spent doing hobbies or in active leisure activities (artistic activities as painting or drawing; doing puzzles or construction games, etc...) is positively associated with the academic achievement (McHale et al., 2001), as well the participation in extracurricular activities programs organized by the school – including sports, clubs and artistic activities (Cosden et al., 2004, Holland & André, 1987). Physical activity is associated with academic performance (Dwyer, Sallis, Blizzard, Lazarus & Dean, 2001). The authors show that “measures of cardiorespiratory endurance, muscular force and power, and physical activity were all related to scholastic ability” (p.235). Similar results are observed by McHale and collaborators (2001) that notice that the 10 years-old children who spend more time in sport activities, are more likely to have better school performance. Several studies emphasize the positive relation between time spent studying or doing homework and the academic achievement (Cosden et al., 2004,). In a trans-cultural study comparing the school performance of the American, and Japanese and Chinese children, Chen and Stevenson (1989), conclude that Chinese and Japanese children have better school results and also spend more time doing homework as well as other schoolrelated activities. The review of the literature that we have presented leads us to the following hypotheses: - 1º ME is positively associated with the children’s academic achievement; - 2º time spent by children Socializing is positively associated with the ME as well as with children's academic achievement, working as a mediator in the association between these two variables; - 3º time spent in outdoor activities, that is, in non-structured activities without adult supervision, is negatively associated with the ME as well as with children's academic achievement, working as a mediator in the association between these two variables; - 4º time spent watching television is negatively associated with the ME as well as with children's academic achievement, working as a mediator in the association between these two variables; - 5º time spent studying and doing homework is positively associated with the ME as well as with children's academic achievement, working as a mediator in the association between these two variables; - 6º time spent reading is positively associated with the ME as well as with children's academic achievement, working as a mediator in the association between these two variables; - 7º the number of activities performed by children is positively associated with the ME as well as with children's academic achievement, working as a mediator in the association between these two variables.

2. Method 2.1 Participants One hundred and thirteen eight-years-old children participated in this study, 53 girls and 60 boys. All children attended the third year of schooling, and were chosen from

9 classes belonging to 8 elementary schools in the Metropolitan Area of Porto-Portugal1. The ME varies between 2 and 17 years, average 9,41 (SD=4.65). Children with more than one year of retention as well having any handicap or significant learning disabilities were not considered. Table 1 presents some demographic information on the participants. From an initial sample of 134 children, 21 were latter excluded by the following reasons: insufficient information (12 children); missing school at the day of the week assigned to complete the time diary, doing that an atypical day (8 children); one child was living in an institution and only went home at the weekend. Table 1: Participants’ characterization N

%

Girls Boys

53

46,9%

60

53,1%

Neighbourhood

Suburban

53

46,9%

Urban 1 to 4 years

60

53,1%

Maternal Education

23

20,9%

5 to 9 years

33

29,2%

10 to 12 years

26

23,0%

13 to 17 years

21

18,6%

10

8,8%

Gender

Missing

2.2 Measures Maternal Education ME was approached in this study as a continuous variable, measured by the number of school years that the mothers completed with success. Academic Achievement To assess children's academic achievement we used the academic competence scale of the Portuguese version of the Social Skills Rating System (teacher form) (SSRS), composed by the following items: 1- Global academic performance; 2- Intellectual functioning; 3- Mathematics aptitude (absolute criterion); 4- Mathematics performance (relative criterion); 5- Reading aptitude/maternal language (absolute criterion); 6- Reading performance/maternal language (relative criterion).

1

In the north of the country, Porto is the second largest city in Portugal: its Metropolitan Area has approximately 1,200,000 inhabitants.

Lemos and Meneses (2002) defined “relative criterion” as corresponding to the “way the student is compared with its classmates” and “absolute criterion” as referring to the performance of the student “relatively to what would be expected for the third year of schooling”. Each item was evaluated in a five-point scale, “1” corresponding to the lower level and “5” to the higher level. Teachers filled out the scale. We decided to carry out the analysis by learning area (absolute criterion in reading and mathematics) as well as considering a global index of school achievement, which results from the average of the six items. Time Use To obtain information about children’s time use we made an adaptation of the time diary from the Child Development Supplement (CDS) of the Panel Study of Income Dynamics from Michigan’s University. Children were asked about their flow of activities over a 24-hour period beginning at midnight. Children were asked to enumerate the activities carried out in two days: the weekday before the day of the interview and the last Sunday. For each activity children were still requested to indicate: a) what time the activity began and finished; b) where they were; c) who was doing that activity with them; d) who was also in that place but not directly involved in that activity; and f) what else were children doing at the same time. All the time diaries reported to the 24 hours of a day, since the 0h00m up to the 24h00m of that same day. The time diaries were filled out chaining the activities one in the other, that is, the time of the end of an activity coincided with the time of beginning of the following one. So, the sum of the time spent in the activities carried out along one day was 24 hours, zero minutes and zero seconds. The activities referred on time diaries were coded in a four digits system according to the codebook from Child Development Supplement (CDS) to the Panel Study of Income Dynamics of the University of the Michigan (PSID, 2004). The 105 different activities reported by children were grouped, in a first level, in the six categories identified by Yeung and collaborators (2001). These six level-1 categories were divided in the 17 level2 categories that we used in our analyses. Table 2 summarizes the categorization system.

Table 2: Time use activities categorization Level 1

Nível 2

Personal Care Activities

Meals Personal Needs and Care Sport activities Outdoor Leisure Activities Other Active Leisure

Play and Leisure Activities

Electronic games Passive leisure Organizational Activities

Achievementrelated Activities

Attending classes, school if fulltime student. Other achievement activities school related Domestic Tasks

Household activities Obtaining Goods and Services Socializing Social Activities

Religious activities Cultural Events

Other activities

Examples Eating meals at home and away from home Washing, showering, bathing, dressing; getting ready,… Playing soccer, basket, doing gymnastics... Playing outdoors with friends; walking for pleasure, bicycling, Drawing, painting, playing indoors... Playing computer games, playstation, etc… Watching television, listening to music, etc... Attending meetings of volunteer, helping organizations; attending a before or after school club… Excluded breaks and meals Studying, doing homework Meal cleanup, clearing table, in door cleaning and chores, picking up, dusting, making beds… Groceries; supermarket; shopping for food; Personal care; beauty, barber shop; hairdressers, tanning. Visiting with others; Party; wedding reception, At bar; Attending services of a church; praying Attending sports; Movies; Theatre, Museums, art galleries, exhibitions…

Travels Sleeping

The total number of activities referred by children, as well as the average time spent in each one was calculated concerning the day of the week (T1) and the Sunday (T0). The daily average time spent in each category of activity, was calculated using the formula [(T1 X 5) + (T0 X 2)] / 7. 2.3 Procedure Data collection concerning time diaries was performed in the elementary schools, by five interviewers previously trained. All children participating in the study had permission from their parents. Each child filled two diaries, one concerning a day of the week (the previous day) and the other a weekend day (the last Sunday). The interview lasted about 30 to 35 minutes. The teachers were requested to fill a questionnaire concerning demographic information as well as the SSRS for each child.

2.4 Results The descriptive data reveals an average of years of maternal education of about 9 years (SD=4.65), in a range from 2 to 17. The academic achievement indicator, evaluated in a scale of 1 the 5, presents the average value of 3.50 (SD =.96). From de several dimensions of children's time use, we can distinguish the more than two hours children spend in Outdoor Play activities in a weekday (M=127.5m; SD =109.8) and the average of about 3 hours that they spend watching television on Sunday (M=180.8m; SD =164.8). Children spend only an average of 1,33m (SD=6.44) per day reading. Being this activity of very low frequency, this is an estimate very underestimated (as we can see in the discussion). In what concerns to the number of activities, we made a t test of repeated measures to evaluate the difference between the number of activities reported on the week and on Sunday. A significant decrease from the number of activities reported on the week (M=23.08, SD=4.59) to the number on reported for the Sunday (M=17.50, SD=4.55), t(112)=12.04, p=.00 is verified. The value eta squared (.56) indicates a strong effect size. It gives the idea that on Sunday children spend more time, doing the same. We find a significant, strong, and positive association (r=.54, p <.01) between the independent variable (ME) and the dependent variable (Academic Achievement). Established the relation between the predictor and the dependent variable, we can now say that a third variable functions like mediator if itself contributes for the relation verified (Baron & Kenny, 1986). To assess the mediator effect of the several dimensions of time use in this relation were made preliminarily the correlations of the categories of time use with the several dimensions of the children's Academic Achievement and with their ME. In the table 3 is presented those correlations, as well the average and the standard deviation of the several variables.

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 10

Table 3. Mean, Standard Deviation and Correlations for the measures (N=113) 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

1. Maternal Education 2. Academic Achievement 3. Global School Performance 4. Portuguese Language _ peers 5. Mathematics _ peers 6. Portuguese Language _ third year 7. Mathematics _ third year 8. Intellectual functioning 9. Socializing _ weekday 10. Socializing _ Sunday 11. Socializing _ weekly average 12. Outdoor Play Activities _ weekday 13. Outdoor Play Activities _ Sunday 14. Outdoor Play Activities _ weekly average 15. Watching television _ weekday 16. Watching television _ Sunday 17. Watching television _ weekly average 18. Studying and doing homework _ weekday 19. Studying and doing homework _ Sunday 20. Studying and doing homework _ weekly average 21. Reading _ weekday 22. Reading _ Sunday 23. Reading _ weekly average 24. Number of activities _weekday 25. Number of activities _Sunday 26. Number of activities _weekly average

1,00 .55** .50** .50** .46** .58** .51** .55** .10 .29** .22* -.03 -.31** -.17 .02 .21* .14 -.03 .14 .00 .21* .10 .22* .23* .13 .23*

1,00 .96** .95** .94** .94** .96** .95** .08 .36** .23* .04 -.21* -.06 -.02 .13 .06 -.05 .07 -.03 .18 -.01 .15 .19* .15 .21*

1,00 .93** .89** .89** .88** .88** .06 .34** .20* -.01 -.23* -.12 .02 .10 .07 -.04 .11 -.02 .17 -.02 .15 .14 .16 .17

1,00 .83** .93** .84** .87** .09 .36** .24** .03 -.19* -.07 -.05 .10 .02 -.04 .08 -.03 .18 -.03 .15 .17 .12 .18

1,00 .81** .92** .88** .00 .29** .13 .03 -.19* -.06 .02 .10 .07 -.01 .04 .00 .11 -.02 .09 .15 .17 .18

1,00 .86** .88** .01 .35** .26** .07 .19* -.03 -.05 .15 .05 -.07 .06 -.06 .22* .04 .21* .21* .17 .23*

1,00 .93** .10 .37** .25** .06 -.21* -.04 .00 .15 .08 -.05 .08 -.03 .13 .00 .11 .23* .12 .22*

1,00 .06 .33** .20* .05 -.20* -.05 -.05 .16 .05 -.07 .05 -.06 .21* -.03 .17 .19* .14 .20*

1,00 .08 .89** -.16 -.12 -.19* -.17 .09 -.08 -.01 .06 .00 -.05 .08 -.01 -.10 .00 -.08

1,00 .52** -.08 -.27** -.19* -.03 .02 -.01 .03 .08 .05 .11 .01 .10 -.16 .11 .17

1,00 -.17 -.23** -.25** -.16 .08 -.07 .01 .09 .03 .01 .07 .03 -.01 .05 .01

1,00 .08 .89** -.18* .07 -.10 -.15 -.04 -.16 -.11 -.03 -.11 .06 -.04 .04

1,00 .53** .05 -.22* -.09 -.02 -.15 -.05 -.11 -.07 -.12 -.11 -.10 -.12

1,00 -.13 -.05 -.13 -.14 -.10 -.16 -.14 -.06 -.15 .00 -.08 -.03

1,00 .15 .83** -.13 -.04 -.14 -.07 .12 -.02 .01 .11 .04

1,00 .67** -.13 -.02 .13 .01 -.07 -.02 .26** .01 .21*

1,00 -.02 -.04 -.03 -.05 .05 -.02 .15 .08 .15

1,00 -.09 .98** -.10 -.06 -.11 .20 .05 .18

1,00 .13 .08 -.03 .06 -.10 .03 -.07

1,00 -.09 -.07 -.10 .18 .05 .16

1,00 .14 .93** -.04 -.01 -.04

1,00 .50** -.04 .04 -.02

1,00 -.05 .01 -.04

1,00 .39** 1,00 .95** .66** 1,00

M SD

9.41 4.65

3.50 0.96

3.37 0.96

3.45 0.98

3.42 1.02

3.58 1.06

3.52 1.02

3.66 1.03

14.46 38.43 21.31 127.5 106.4 121.4 84.30 180.8 111.8 40.73 8.01 31.38 1.33 53.34 70.37 44.49 109.8 150.0 92.38 88.40 164.8 84.28 43.39 24.01 31.13 6.44

1.33 6.81

1.33 5.25

23.09 17.50 21.49 4.59 4.55 3.84

**p< .01; *p< .05

25.

26.

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 11

The results presented confirm the association between ME and the several dimensions of children's academic achievement, being the strongest correlation between ME and the performance in the Portuguese Language. Meanwhile we tried to identify which the time use variables that fulfilled the two first necessary conditions to confirm de mediation's hypotheses: “(a) variations in levels of the independent variable significantly account for variations in the presumed mediator, (b) variations in the mediator significantly account for variations in the dependent variable” (Baron & Kenny, 1986, p. 1176). Two categories of activities arise, both occurred on Sunday, who are related so much with the ME as with the global index of academic achievement: with a positive correlation time spent Socializing, and with a negative relation time spent in Outdoor Leisure Activities. The infants, whose mothers have more years of education, are also those with better school results, and those that, on Sunday, spend more time in Visits and Social Events, and less time in Outdoor Leisure Activities. Time spent watching television does not present significant correlation with any indicator of children's academic achievement; regarding the ME, presents a significant and light correlation (p.<.05) in time spent watching television barely on Sunday. Time spent studying or doing homework did not present any significant correlation with the indicators of children's academic achievement, or with the ME. Time spent reading to the week fulfilled the two first conditions for mediation, presenting a significant and light correlation (p.<.05) with the ME and with the performance in the Portuguese Language. Also time spent reading to the week we can observe a light and significant correlation (p.<.05) with the index “intellectual functioning” reported by teachers. However is important to notice that, of the 113 participants, only 5 referred to read, being for that the average a very reduced value. The number of activities carried out (so much to the week, as on average weekly, although be perceived that the effect is especially by what happens to the week) revealed a positive relation with the ME, with the global index academic achievement, with the indicator of achievement in the Portuguese Language and in Mathematics and with the intellectual functioning. Known these results, we can continue to test the model of mediation with the following variables: time spent Socializing on Sunday and time in Outdoor Leisure Activities on Sunday, time spent reading and the number of activities that infants reported in their time diaries (in these two last cases was considered the week time). The Figure 1 presents the values of linear regression between these variables represented according to the model proposed by Baron and Kenny (1986, p.1176).

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 12

Figure 1: Linear regression values between variables according to the mediation model Mediators Time Use Variables

Socializing _ Sunday Outdoor Leisure .2 9* * β=

Independent Variable

β=

**

Number of activities _ weekday

* .23 β=

.21 *

β= .1 9

*

Academic Achievement

β=.55**

ME ME

Portuguese Language

β=.58**

β=.2 1*

* β=.22

Dependent Variable

Predictor

.31 β=

β= .3 6* *

Activities _ Sunday

Reading _ weekday

It is important now to test the third condition for that we could speak on mediation: regressing together the independent variable and the mediator in the dependent variable, the effect of the independent variable in the dependent variable is diminished (Baron & Kenny, 1986). Table 4: Regressing Academic Achievement with Maternal Education and Time Use variables β

SE

t

R2

Maternal Education Number of Activities _ weekday

.52** .11

.11 .12

6.10 1.26

.30

Maternal Education Outdoor Leisure Activities _ Sunday

.53** -.05

1.32 .00

6.01 -.59

.29

Maternal Education Socializing _ Sunday

.47** .26**

.11 .00

5.63 3.10

.35

**p<0,01; *p<0,05

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 13

Table 5: Regressing Portuguese Language Performance with Maternal Education and time spent reading _ weekday

Maternal Education Reading _ weekday

β

SE

t

R2

.55** .10

.02 .00

6.75 1.04

.34

**p<0,01; *p<0,05 We can observe an evidence of mediation regarding the variable Socializing on Sunday. In the regression analysis putting together the predictor (ME) and the mediator (Socializing on Sunday), the effect of the ME in academic achievement, ß = .47, p<.01 is lower than in the straight relation performed, ß = .55, p <.01. In this equation the mediator continues to affect the dependent variable, ß = .26, p <.01. For a perfect mediation this association bust be zero or no longer significant. However, as Baron and Kenny (1986) says “because most areas of psychology treat phenomena that have multiple causes, a more realistic goal may to seek mediators that significantly decrease than eliminating the relation between the independent and dependent variable” (p.1176). In figure 2 there is a summary of the model of mediation of this variable represented according the proposal of Baron and Kenny (1986, p.1176).

Figure 2: Complete mediation model for the mediator Socializing _ sunday Mediator

Socializing _ Sunday

Independent Variable

.3

6* *

β=.26* *

β=.47* *

ME β=.55**

Academic Achievement

Dependent Variable

Predictor

β= .2 9

**

β=

Regarding the remaining variables (Outdoor Leisure Activities on Sunday, Number of activities and time spent reading in a weekday), despite of a light decrease in the effect of the ME in the equation where this variables appear as mediators, we can't to speak of mediation because, in that case, these variables don’t present now a significant effect in the dependent variable.

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 14

3 Discussion The main goal of this study was to analyse the relation between ME and the children's academic achievement, trying to perceive in what measure the way children uses their time has a mediator effect in this relation. In first place, data confirms the association between ME and children's academic achievement. The increase of social life in families where mother has higher levels of education, referred by Carvalho (2001) can be seen in the results of this study, in the significant and positive correlation between ME and the time spent Socializing on Sunday. We can also verify the association of this variable with children's academic achievement. Of the several dimensions of children's Time Use, time spent Socializing on Sunday – go to a bar, go to a birthday party, go home of relative or friends – is the unique who functions as mediator in this association between ME and children's academic achievement. This dimension seems to be important for children's school performance. His effect can be related with a higher range of experiences and a contact with more persons and different contexts, which translate a more active social life. It could also be important here to perceive the effects of the social context of interaction. For example, we can be put the hypothesis that the effects verified in this category will be also related to the fact that these activities habitually will be carried out in family and/or with adult’s supervision. Time spent in Outdoor Leisure Activities did not arise as a mediator variable in the relation between ME and children's academic achievement. Perhaps due the epoch of the year in that data were collected (with sun and nearby the end of the academic year) we couldn’t find enough variability between the subjects to speak of mediation. However, is evident the significant and negative association between time spent in Outdoor Leisure Activities on Sunday and ME by a side and the several dimensions of the school performance by another. Parke and Buriel (1998) refer the negative consequences of the decrease supervision and parental monitoring in the cases of children with delinquent and anti-social behaviours. These data, reinforce the idea of Hofferth and Sandberg (2001) that is desirable an equilibrium between activities carried out at home and activities carried out outside of home, as well the need of adult's supervision as mediator of the relations with peers. Not consistently with the literature (Hofferth & Sandberg, 2001; Lopes and Rabbit 2002; Mc Hale et al., 2001) the time spent watching television, particularly time spent watching television on Sunday, appears positively associated with ME. We can also verify a negative correlation of this variable with the time spent in Outdoor Leisure Activities, also on Sunday (and from this variable also a negative correlation with ME), giving the sensation that, children of families with higher levels of ME have tendency to spend less time playing outdoors and more time at home watching television. Time spent watching television also doesn't present any negative correlation with children's academic achievement, not according with the reduction hypothesis proposed by Shin (2004). Another result contrasting with what is verified in the literature is the absence of any significant correlation between time spent studying or doing homework, so much with the several indicator of the academic achievement, as with ME. With the school year nearby his end in the data collection period, and with the question of evaluation in many certain cases already resolved, it can be happened that for many of the children studying or doing homework wasn’t necessary at that time. In fact, is little the time spent studying or doing homework referred by children in this study. Using the same methodology, Hofferth and Sandberg (2001) counted a medium weekly of 44 minutes daily while, in this I study, we found an average of 31 minutes daily studying or doing homework (40 minutes in day of the week and, on weekend, practically nonexistent). On the other hand, there is a great

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 15

variance in the results found, having so many children that referred not to have spent any time studying or doing homework as one child who referred to have spent more than three hours in this activity. Perhaps in a study with data collection made in the middle of school year would be possible to find different results and then find the expected correlation. We found the expected positive correlation of time spent reading with academic achievement in Portuguese Language - reading seems to be an important indicator for better results in this area. Being so little the time spent by these children reading, (even the few who referred to read), and knowing that the Portuguese have the lower index of reading in the European Union (APEL, 2004), we can find here a plausible explanation for the difficulties in the Portuguese Language, and great percentage of illiteracy verified in our country. However, in this study, time spent reading is not a mediator between ME and academic achievement in Portuguese Language. Perhaps this little effect could be related with the low frequency (only 5 in a total of 113) of children who referred read for leisure. This small average value that is verified is a indicator of one of the main limitations of the time diaries – as referred by Hofferth and Sandberg (2001) this instrument is not a good measure for low frequency activities since it supplies underestimated results. It is also possible to verify the importance of the number of activities referred by children. A great number of activities seem to indicate a richer fan of experiences and then the positive association with the academic achievement. Naturally, given the correlacional nature of the analyses, it is also possible to put the hypothesis that children with a better school performance could also be those who, in the interview, are going to better discriminate their daily activities and, because of that, refer a great number. Being also positively associated with ME, is not confirmed however the effect mediator in the relation between this variable and the academic achievement. The presented results seem to alert for the importance of the weekend and the leisure time in the children’s development. Although considered period of "rest", what the child makes in the weekend seems to have an important influence particularly in its academic achievement. The data also seem to confirm the evidence that the parents' level of education is one of the main factors to contribute for qualitative alterations of children's time use (Carvalho, 2001) and direct or indirectly for an improvement of children's academic achievement. However it's important to develop more studies for a better understanding of this association. It could be important for example a better understanding of the influence of children's characteristics in all these dimensions - the "child effects" described by Russel & Russel (1992). For example, as referred by McHale and collaborators (2001) “children who are well adjusted are better equipped to immerse themselves in the kinds of constructive activities and social contexts that support adaptive functioning” (p. 1775). It could be important the development of the mediation model, an investigation area who needs more studies and theories (Russel & Russel, 1992), using for example methodologies like the “path analysis” or the “Structural Equation Modelling”. Several others time use measures, but also children, family and environmental measures (namely from the school), could be used to better explain this path, certainly not linear, between ME and children’s academic achievement. In fact, the analysis made in this study reported only univariated effects. We believe that children’s time use has a multivariate effect. Because of that, it could be relevant to analyse several dimensions altogether to understand for example how they are grouped in routine profiles. It could be also important to use a Socioeconomic status (SES) in spite off only this ME indicator. It could be possible for example to use a composite measure (and compare the results with these with the ME indicator) or use the concept of “Family Capital” proposed by Marjoribanks (2001) trying to understand if the results will go in the same

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 16

way as with ME. For example, in what concerns specifically academic achievement, the family environments’ assessments using the “Home Observation Measurement of the Environment (HOME) are better predictors of children’s academic achievement than the maternal education” (Carter & Murdock, 2001, p. 4). The use of more objective measures of school performance, for example performance tasks, would permit to control a role of variables proper of pertaining to school context when the criterion of evaluation is only the judgment of the teacher. It could also be relevant to replicate this study in another season. The data of McHale and collaborators (2001) show for example that the children spent more time watching television during the months of Winter and that its decrease in the heat months is related with higher levels of involvement in sport activities and outdoor play activities in the months of Spring and Summer. They detach thus the "temporal rhythms" inherent in children's activities (McHale et al., 2001, p 1774). The data of this study could be crossed with one its rejoinder with data collected for example in November/December with the existence of a new variable of analysis: season of the year. Finally, it seems important to develop in our country the investigation on the children's time use and its implication in the development. Today we attend the development of the children observing great handicaps in their experiences, or without making the idea of the developmental consequences of what they are doing. For example, with the proliferation of computer use and the Internet for multiple goals, occupying its use one great parcel of children's time, what it will result in the future generations? Studies as of Subrahmanyan, Greenfield, Krauf and Gross (2001) show what it can be done. For now we only can make inferences about results obtained in another realities without transcultural studies.

4 References Baron, R.M. & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. Bronfenbrenner, U. (1979). The Ecology of Human Development: Experiments by Nature and Design. Cambridge, Mass.: Harvard University Press. Caldas, S.J. & Bankston, C. (1997). Effect of school population socioeconomic status on individual academic achievement. Journal of Educational Research, 90, 269-277. Cappella, E., & Weinstein, R. (2001). Turning around reading achievement: Predictors of high school students’ academic resilience. Journal of Educational Psychology, 93, 758–771. Carter, A.S. & Murdock, K.K. (2001). The Family as a Context of Psychological Functioning. In E.L. Grigorenko & R.J. Sternberg (Eds.). Family Environment and Intellectual Functioning (1-23). Lawrence Erlbaum Associates. Mahwah, New Jersey. Carvalho, M. J. (2001). Gender and Children’s Time Use. Rio Grande do Sul-Brasil, Retirado do sítio «http://www.stmarys.ca/partners/iatur/» em Fevereiro de 2004. Chen e Stevenson (1989). Homework: A Cross-cultural Examination. Child Development, 60, 551-561. Cosden, M., Morrison, G., Gutierrez, L. & Brown, M. (2004). The Effects of Homework Programs and After-School Activities on School Success. Theory Into Practice, 43(3), 220-226. Crane, J. (1996). Effects os Home Environment, SES, and Maternal Test Scores on Mathematics Achievement. The Journal of Educational Research, 89 (5), 305-314. Dwyer, T., Sallis, J.F., Blizzard, L., Lazarus, R. & Dean, K. (2001). Relation of Academic Performance to Physical Activity and Fitness in Children. Pediatric Exercice Science, 13, 225-237.

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 17 Gilliam, B., Gerla, J., & Wright, G. (2004). Providing Minority Parents with Relevant Literacy Activities For Their Children. Reading Improvement. 41(4), 226-235. Hoff, E., Laursen, B. & Tardif, T. (2002). Socioeconomic Status and Parenting. In M.H. Bornstein (Ed.). Handbook of Parenting. (vol. 2). Lawrence Erlbaum Associates, Publishers. Londres. Hofferth, S. & Sandberg, J. F. (2001). How American children spend their time. Journal of Marriage and Family. 63(2), 295-308. Holland, A. & Andre, T. (1987). Participation in Extracurricular Activities in Secondary School: What is Known, What Needs to Be Known? Review of Educational Research, 57(5), 437-466. Larson, R. & Verma, S. (1999). How Children and Adolescents Spend Time Across the World: Work, Play , and Developmental Opportunities. Psychological Bulletim, 125 (6), 701-736. Lemos, M.S. & Meneses, H.I. (2002). A Avaliação da Competência Social: Versão Portuguesa da Forma para Professores do SSRS. Psicologia: Teoria e Pesquisa. 18 (3), 267-274. Lopes, M. G. C., & Coelho, E. (2002). Diferenças e Semelhanças entre o Uso do Tempo das Crianças e dos Adultos em Portugal. Documento apresentado na Annual Conference 2002 15 - 18 October Lisbon, Portugal da International Association of Time Use Research (IATUR), retirado do sítio «http://pascal.iseg.utl.pt/~cisep/IATUR/time%20conference%202002.htm» em 12 de Janeiro de 2004. Ma, X. (2000). Socioeconomic Gaps in Academic Achievement Within Schools: Are They Consistent Across Subject Áreas?. Educational Research and Evaluation, 6(4), 337-355. Marjoribanks, K. (2001). Family Capital and Cognitive Performance. In E.L. Grigorenko & R.J. Sternberg (Eds.). Family Environment and Intellectual Functioning (49-70). Lawrence Erlbaum Associates. Mahwah, New Jersey. Mc Hale, Crouter & Tucker (2001). Free-Time Activities in Middle Childhood: Links with Adjustment in Early Adolescence. Child Development, 72 (6), 1764-1778. Okpala, C.O., Okpala, A.O. & Smith, F.E. (2001). Parental Involvement, Instructional Expenditures, Family Socioeconomic Attributes, and Student Achievement. The Journal of Educational Research, 95(2), 110-115. Palacios, J. Hidalgo, M. V. & Moreno, M. C. (2001). Familia Y Vida Cotidiana. Familia Y Desarrollo Humano. (70-89). Alianza Editorial. Parke, R. D., & Buriel, R. B. (1998). Socialization in the family: Ethnic and ecological perspectives. In N. Eisenberg (Ed.). Handbook of child psychology (5th ed., pp. 463-552). New York: John Wiley. PSID (2004). Home page of the Panel Study of Income Dynamics, taken from the site http://psidonline.isr.umich.edu/, in 12 de February, 2004. Robinson, J., (1988) Time-Diary Evidence About the Social Psychology of Everyday Life. In McGrath (editor). The Social Psychology of Time – New Perspectives (135-148). Sage Publications. Russel, A. & Russel, G. (1992). Child Effects in Socialization Research: Some conceptual and data analysis issues. Social Development, 1(2), 163-184. Sheldon, S. B. (2002). Parents’ social networks and beliefs as predictors of parental involvement. The Elementary School Journal; Chicago; Mar.2002. Shin, N. (2004). Exploring Pathways From Television Viewing to Academic Achievement in School Age Children. The Journal of Genetic Psychology, 165(4), 367-381. Subrahmanyan, K., Greenfield, P., Krauf, R. & Gross, E. (2001). The impact of computer use on children’s and adolescents’ development. Applied Developmental Psychology, 22. 7-30. Walker, S. O., Petrill, S. A. & Plomin. R. (2005). A Genetically Sensitive Investigation of the Effects of the School Environment and Socio-Economic Status on Academic Achievement in Seven-Year-Olds. Educational Psychology,25 (1), 55–73. White, K.R. (1982). The relation between socio-economic status and academic achievement. Psychological Bulletim, 91, 461-481.

Socioeconomic status, academic achievement and time use in eight year old Portuguese Children 18 Wright, C., Diener, M. & Kay, S.C. (2000). School readiness of low-income children at risk for school failure. Journal of Children & Poverty, 6(2), 99-117. Yeung, W. J., Sandberg, J. F., Davis-Kean, P. E., & Hofferth, S. L. (2001). Children’s time with fathers in intact families. Journal of Marriage and Family. Feb. 2001, 136-154. Zady, M. & Portes, P. (2001). When Low-SES Parents Cannot Assist Their Children in Solving Science Problems. Journal of Education for Children Placed at Risk, 6(3), 215–229.