Diet & Welfare Policies

DIET


Effect Size d= 0.12  (Hattie's Rank=123)

One meta-analysis was used:
Karvale, K. & Forness,S., (1983) Hyperactivity and diet treatment. Journal of Learning Disabilities. 16(6), 324-330.

This paper is not measuring diet as it relates to improving student achievement. But rather diet modification as a treatment of hyperactivity. In addition, the research is only about a particular type of diet modification based on the hypothesis that, "the ingestion of artificial (synthetic) food additives (colours and flavours) and naturally occurring salicylates in foods results in hyperactivity and learning disabilities in children... It was suggested that treatment be based on the Feingold Kaiser-Permanente (K-P) diet which is designed to eliminate all foods containing natural salicylates and artificial food additives from the diet" (p324).

Hyperactivity was measured by a number of different tests: e.g., Connors' Scale parents, Connors' Scale teachers, attention, disruptive behaviour, impulsiveness, global improvement, learning ability and hyperkinesis. The researchers report (p327) a summary of effect sizes as listed below:


Hattie seems to have averaged all the categories except for 'Impulsivity' and reports an average, d = 0.12. There is no explanation about this in his commentary on the diet on pages 52-53 of VL.

So once again Hattie includes items that are not measuring achievement. So significant doubts are raised about the validity of this analysis in the context of VL - i.e., improving student achievement.

From docendo:
"There is one meta-analysis, Kavale and Forness (1983). I can only access the abstract but it’s clear that despite the missing clause in Hattie’s summary, the meaning that I had assumed he intended does match this meta-analysis. Equally, it is clear that this is very specifically looking at ADHD and not children without this diagnosis. Essentially this paper states that the studies analysed do not provide evidence to support the earlier hypothesis that dietary changes could have a positive effect on ADHD symptoms. I’m guessing that the outcome measure was not academic achievement, but more likely some behavioural measure, which reminds me again that Hattie seems rather blasé about what his meta-analyses are measuring."



WELFARE POLICIES


Effect Size d= -0.12  (Hattie's Rank=135)


Hattie used 1 meta-analysis which seemed to give a negative effect size of -0.12. Note the CLE probability statistic of -8% is wrong as probabilities cannot be negative.

The question should be asked, what were the specific welfare policies?

Schools use a wide range of policies from money given to families to help buy books, food, etc to 1-1 social worker and psychologist help, to the addition of certain materials in the curriculum, e.g., the PERMA model and social/emotional learning, to assigning specific teachers to watch over particular students to act as mentors.

Many schools interpret the above result as welfare policies, IN GENERAL, do not improve student achievement. Once again this needs to be looked at in greater detail.

Hattie's only comment (p65) on this influence is that the 1- meta-analysis used found,

'close to zero effects from students in families who received welfare compared to those not receiving welfare. While they make much of an effect size of d = –0.10, by claiming that the effects on adolescents were “significantly worse”, it is difficult to imagine the visible effects of findings such as about four percent fewer mothers in the welfare program group reporting that their child performed above average, and only about two percent more of this group of mothers indicating that their child repeated a grade. There are certainly many other effects of welfare programs for these families that are beneficial, but it seems that there are other more powerful effects on achievement than the welfare status of the family.'

WHAT WERE THE WELFARE POLICIES?


The policies investigated were ones which were 'designed to increase parental employment' (p404).

The key policies considered in this literature are (p401):

(1) work mandates, which require parents to work or to participate in education and training programs designed to enhance their employability; 

(2) financial incentives, such as enhanced earned income disregards that allow working welfare recipients to keep more of their welfare benefits than under Aid to Families with Dependent Children (AFDC) or earnings supplements that are contingent on 30 hr or more of work per week; and 

(3) time limits, which limit the length of time families can receive cash assistance. These policies are designed to increase maternal employment and family income, and reduce receipt of public assistance.

HOW WAS STUDENT ACHIEVEMENT MEASURED?


The achievement was measured at around the 60 month follow up (p405).

'This outcome was measured using maternal responses to the following question: 'Based on your knowledge of your child's schoolwork, including report cards, how has your child been doing in school overall?” Mothers expressed their responses on a 5-point scale ranging from 1 (not well at all) to 5 (very well)' (p406).

The authors concede,

'Maternal reports of children's school performance are less than ideal' (p407).

RESULTS




Hattie appears to report the wrong effect size -0.12, the table shows it is -0.11. Whilst a small error this just adds to the many other errors Hattie makes.

In their discussion (p413), the authors try to explain why the results are negative. Two issues seem to emerge: there is less parental supervion of homework, etc. and adolescents take on the parenting role of younger siblings leaving less time for studies.

In any case, this meta-analysis does not appear to be relevant to the many varied and different welfare policies schools use.