Change Valve In Cooker

year Stupidity Inequities several PISA

After reading the press release embarrassing Ministry and listen to the string of nonsense words that have made the poor Beder in PISA press conference - where was the minister, by the way? - I decided to write something about it. I'm used to that the Ministry does not addresses the data, which blatantly lie when it suits him, to manipulate the interpretation of the data to justify its policy, etc. There would be no big deal if it were not the extreme conservatism of the PSOE in education since 1990 is hurting our country is extremely serious.

For example, Beder, saying in a press conference that the best vaccine against school failure is the early schooling, and that slightly less will not stop until schooling for all children under three years. PISA does not say anything like that (although most do not leave school early have very good standing), while a OECD report on the English education itself clearly says that we miss the nursery. But as the Pisuerga passing through Valladolid, loosely as a reference to the sole indicator with positive evolution at my disposal, so I'm happy. Another public policy based on a belief rather than a fact.

But this is not what I wanted to talk, after all, it's just money thrown away, but the concept of equity that the Ministry manages the press release, an ideological concept which serves to avoid seeing reality and to continue the inaction that characterized him.

As I explained once , the Ministry manages a concept of equity based on who has few statistical differences among students, which technically corresponds to the variance or standard deviation. Is a translation of an economic concept -An economic good is, By definition, a scarce commodity, not an economic one, as the educational level: the more knowledge one has is independent of that other have, because knowledge is not exhausted.

For example, our educational system virtually outlaws excellence: from percentile 75 (ie, if we compare the results of 25% of the best students from each country) and we are the last country in Europe, while our 95 percentile borders on the ridiculous. If we go "is a simple simulation, a 2% reading scores of students between 75 and 80 percentile, 4% of students between 80 and 85, and so up to 10% the 95 percentile, the average of Spain would go eight points, but the standard deviation would be 100, ie the same as the OECD average. That is, if we go up a bit the results of the best students, something that the education system in almost any European, seems capable of achieving the improvement is small, but equity is going to hell. And that without changing the value of 75% of the students left. Why? For our famous statistic is only fair that effect our system is unable to form bright students.

As I said at some point yesterday, I do not know where or to whom, if we really believe we are in the knowledge society, and that bright students are the largest producers of wealth in the morning, our lack of excellent students today is just a number. But in twenty years what we call crisis.

Actually, the issue of equity is no example Spain, much as it says the Ministry. Educational equity consists of three elements: first, that no student falls below a minimum, and second that the circumstances of the students (especially those that can not be changed: sex, race, place of birth, mother tongue ...) No seriously affect in educational outcomes, and third that a student with ability and interest to get to where he want. In the first case we do regularly in the third very bad, as we have seen. We must now examine one aspect of the second. As the Ministry in the press release:

"As for the Autonomous Communities, PISA shows that the results between the two is minimal, only 4%, ie establishing that the English education system is characterized by a large equity. "

confess that I understand what has cost 4%, and a few friends who know of it too. At first I thought they were related to effect size, but miscalculated. Then I found suggests that this refers to the percentage of variance explained by the fact that a Students live in one region or another. Which I think is very good, but it is a fact that has little to do with equity.

I'll start the argument in the style of the Cheshire Cat: To begin with, nobody in their right mind would say that the results from Finland and Spain are "fair" in Finland is 536 points and Spain in the 481, Finland August 1 % of students below Level 2 and Spain 19%, etc. How do we know? First, because the differences between the two countries are significant (ie there are not due to chance). But, with a handle samples as large as PISA, it is easy for a difference is significant. PISA has therefore begun to measure so-called effect size (http://en.wikipedia.org/wiki/Effect_size), ie if the difference is not significant, but whether it is substantive. And, if it is, the magnitude of the substantiation.

himself PISA ( Volume II, pg. 148) explains how to calculate the effect size (using Cohen's delta ) and the interpretation of values, not substantive to values \u200b\u200bbelow 0.20 , unimportant between 0.20 and 0.50, a medium effect between 0.50 and 0.80, and most importantly above 0.80. However, the size of effect between the results from Finland and Spain in the scale reading is 0.63.

Let the English regions. In the table I placed the size of the effect of the differences between English regions. In gray are the non-substantive, paler blue small effects, effects in bright blue and red means large effects.

size effect on the reading scale. PISA 2008
ARA	-0.391
AST	-0,317	0,056
BAL	0,044	0,429	0,355
PVA	-0,381	0,012	-0,045	-0,420
CAN	0,144	0,531	0,452	0,098	0,522
CNT	-0.305	0.081	0.022	-0.344	0.070	-0.444
CYL	-0.483	-0.094	-0,145	-0,519	-0,107	-0,621	-0,173
CAT	-0,432	-0,036	-0,091	-0,470	-0,048	-0,574	-0,118	0,060
CyM	0506	0874	0787	0458	0867	0367	0789	0958	0918
GAL	-0.284		0.105 0.044 -0.324			0.094 -0.424 0.023			0.198 0.142 -0.772
RIO	-0.411	-0.034 -0.086		-0.448 -0.046		-0,546	-0,112	0,057	0,000	-0,880	-0,135
MAD	-0,483	-0,094	-0,145	-0,519	-0,107	-0,621	-0,173	0,000	-0,060	-0,958	-0,198	-0,057
MUR	-0,225	0,182	0,115	-0,267	0,171	-0,371	0,095	0,279	0,222	-0,733	0,072	0,210	0,279
NAV	-0,416	-0,024	-0,079	-0,454	-0,036	-0,556	-0,105	0,071	0,012	-0,899	-0,129	0,011	0,071	-0,207
	AND		ARA AST BAL		PVA	CAN	CNT	CyL	CAT	CyM		GAL RIO	MAD MUR
Source: Compiled from data from PISA 2009.

Let's dispense with Ceuta and Melilla (CyM), with abysmal distances (the Ministry should answer for these data because they are territories under its direct management, but has preferred to go for equities). However, differences between Castile and León and Madrid to the Canary Islands (CAN, Cantabria CNT) have a size of 0.62. Well, one hundredth less than differences between Finland and Spain. Moreover, if we do not consider Finland, the differences between European countries outside the OECD are similar, if we consider the effect size (and METICULOUSLY differences) that exist between the English regions.

is, in a country like Spain, with the same education system, same teachers, same allocation to schools, and cultural homogeneity, linguistic, socioeconomic, etc. much higher than in Europe, is that we have some differences between regions similar to those between European countries, the same between the Netherlands and Austria or the Czech Republic between Castile and Leon (or Madrid) and the Canaries.

already knew that being born in one country or another could make the course a student's education. It seems quite clear that also born in a English region or another. This is what the Ministry considers minor differences.

ADDED:

In a commentary calling this table:

Communities RIO 280 VAL

language test for Spain,
	ESP	CAT	GAL	VAS	VAL	MISS	Total
AND	1.411	0	0	0	0	5	1.416
ARA	1.506	0	0	0	0	8	1.514
AST	1.535	0	0	0	0	1	1.536
BAL	259	1.201	0	0	0	3	1.463
CAN	1.446	0	0	0	0	2	1.448
CAT	3	1.369	0	0	0	9	1.381
CLM	239	0	0	0	0	0	239
CNT	1,516	0	0	0	0	0	1,516
CYL	1,510	0	0	0	0	5	1,515
CYM	1,363	0	0	0	0	7	1,370
EXT	134	0	0	0	0	0	134
GAL	45	0	1,540		0 0 0		1585
MAD	1449		0 0 0			4 0 1453
WALL	1320	0		0 0 0		1	1321
NAV	1079	0 0			425 0 0		1504
PVA	4022	0		743 0 0		3	4768
		1,284 0 0			0 0 4		1,288
			0 0 0		156	0	436
	Total 20,401 2,570			1,540 1,168 156		52
25,887 SOURCE: Compiled from data from PISA 2009.

not entered as a comment and I chose to put it here. The figures represent the number of students evaluated in each language.