On Schools and Survival

TJ Palanca
4 min readSep 11, 2018

A look at dropout rates in the Philippines (EduData Part 1)

This article is part the EduData series, which explores the state of the Philippine education system, and what we could do to make it better.

EduData Series

Education is one of the cornerstones of development, particularly in a country where majority of the population is of school age. Luckily, the Department of Education has released tons of data that allows us to shed light on the state of the Philippine education system. For the first post in this series, we will focus on dropouts and where, grade level-wise and geographically, they occur.

Swimming against the tide

What percentage of students that enter Grade 1 are likely to graduate high school? We first compute the survival and dropout rates over time. The cumulative survival rate (top panel) is the proportion of the Grade 1 class that is expected to graduate high school. The single-year dropout rate (bottom panel) is the attrition of the class for that particular grade level.

By Time

SWIMMING AGAINST THE TIDE — Cumulative survival rates over grade levels

There are some things we can say about the data:

  • Survival rates have been improving steadily over the past three year, with about 60% of the initial Grade 1 class graduating high school in 2015, compared to about 50% in 2013.
  • The survival rate improvement can be attributable to the Grade 2 dropout rate. More and more students are choosing to proceed after Grade 1 in 2015, compared to past years.
  • Once a student has reached Grade 3, the chance that he/she will be unable to graduate elementary school becomes small. However, making the leap from elementary to high school is still difficult.

By Gender

Let’s also flip the facets and compare the cumulative survival rates and dropout rates by gender.

BATTLE OF THE SEXES — Cumulative survival rates by gender, 2013–2015

There are a couple of very interesting findings from this chart:

  • Females have a higher propensity staying in school than males, but fewer are given the opportunity to attend high school after graduating elementary.
  • The gender gaps in dropout rates have remained generally unchanged over time.

If we were to learn from this data, we could make the following recommendations:

  • To get the most impact, focus on ensuring that new Grade 1 students make it through to Grade 2. For females, focus on ensuring a transition from elementary to secondary school.
  • Focus on retention for males throughout the grade levels.
  • While we cannot drill down to the root cause of the issues, educators may have some insight. If you are an educator, feel free to share your thoughts with us!

By Location

Broad averages can conceal a lot of information, and there may be large disparities in dropout rates in different parts of the country. Let’s take a look at the geographic distribution of dropout rates and determine whether there any any ‘hotspots.’

The conflict-torn Zamboanga-ARMM region has always been plagued by survival rates as low as 10%. The situation has not get any better over time. In recent years, however, Eastern Visayas has grown into a significant hotspot, probably due to migration out of the region and exacerbated by the effects of typhoons, especially Yolanda. The effect in Eastern Visayas is more pronounced for males than females. A small hotspot is also developing for males in Central Luzon.

If this is the state of the Philippine education system, what can we suggest?

  • Prioritize these ‘hotspots’ for intervention, determine the causes behind the dropouts in these areas and resolve the issues.
  • Programs such as the Conditional Cash Transfer (CCT) program may have larger impact in these areas in terms of reducing dropouts.
  • This is just the first in the EduData series! In the next few installments, we’ll be focusing on capacities in terms of teachers, rooms, and budget, so stay tuned!

Data, code, and computation requests may be made through the blog contact form.

Links

You might like my other posts

  • On Benford’s Law : Determining import fraud risk using customs data. If you were to think of the first digits of a group of related variables, say, tax payments, it would be intuitive to think that the digits would more or less be evenly distributed between 1 to 9. However, it turns out that the exponential nature of growth makes it such that the first digits are more likely to be 1’s than any other number.
  • On Latin Honors : How hard is it to get to the top? An article from the Inquirer takes a look at how there are more latin honors graduates in Philippines colleges and universities than in years past. We can view comparative data for selected universities to know, more or less, how much weight a *cum laude* title carries for a particular school.
  • On the MRT : A Capacity Conundrum. The Metro Rail Transit Line 3 (MRT) has been operating at 142% capacity since 2004. New prototype trains have been scheduled to arrive in 2015, but the actual deployment will still be in the following year. How bad is the current train situation? Let’s find out through data!

Originally published at www.tjpalanca.com, and also at on GMANews SciTech.

--

--

TJ Palanca

Data-focused CTO at First Circle. Formerly Data Scientist at Uber, Grab, and EY. Fan of aviation, pintxos, and Bayesian statistics.