PAPER TITLE: A GENERAL FRAMEWORK FOR ASSESSINGEDUCATION POLICIES TOPIC: AGENERAL FRAMEWORK FOR ASSESSING EDUCATION POLICIES Abstract: Researchwithin the education sector in India in last two decades has highlighted thepoor and ever worsening situation of quality of education within the Indianprimary education system. Not only do we fare poorly on overall indicators ofeducation such as literacy as compared to other rapidly developing nations, weoccupy the lowest spots across the world when it comes to quality of learningoutcomes.
There has thus been a welcome change in the way we look at educationpolicy in recent years, shifting gears from measures that value quantum such asenrolment ratios to quality-driven metrics such as attendance rates andpedagogical considerations. Recentstudies have focused on a data-driven approach to analyzing educationalpolicies, however the results do not seem to be comprehensive, and each studyfocusses on disparate aspects and metrics for assessment of the situation ofeducation. Thus, the scattered nature of available data impedes one’s abilityto attain an overall picture. Thispaper addresses this very problem of the lack of a general framework forstudying/assessing educational policies. The paper focusses on four aspects -Accessibility, Equity, Quality and Efficiency- deemed as the four essentialpillars of a good education policy after comprehensive literature review. Themethodology for assessing performance on these four metrics has been carefullyoutlined within the paper. Further the effectiveness of some of the existingmeasures has been examined; additionally some modifications have been suggested.As an illustration consider accessibility, while the most widely used parameterto study accessibility is the gross enrolment ratio (GER), the paper not onlyhighlights the effectiveness of GER as a measure but also aims to understandthe importance GER with the implementation of recent policies by discounting itfor factors like attendance.
The other three factors- quality, efficiency andequity- have been examined in a similar fashion. Thepaper further identifies factors affecting each of the four pillars throughextensive literature review. A substantial portion of the paper is devoted tothis very quest, careful regression analysis has been carried out to understandthe impact of underlying factors on each of accessibility, quality, efficiencyand equity. As an illustration, existing data has been used to analyze theimpact on quality of the following- infrastructure (such as proportion ofschools with electricity or playground), nutrition (proportion of schools thatprovide mid-day meal, state-level health indicators), education level ofinstructors, proportion of teacher training institutes and performance ofstudents on mathematical/language ability tests. Additionally, lack of data onsome missing determinants has also been highlighted, such as measures thataccount for pedagogy and their impact on quality. Similar exercises werecarried out for accessibility, efficiency and equity. Withthe above stated frame-work the paper finally aims to understand the overallimpact of various pre-existing policies on primary education, highlighting theachievements and lessons from the policies that have historically beenimplemented to the shortfalls of the most recently implemented policies.
Methodology:Ourmethodology is based on the following ? VARIABLES TO JUDGE THE PARAMETERS? INDEPENDENT VARIABLES DETERMININGTHESE PARAMETERS? REGRESSION ANALYSIS TO CHECKHYPOTHESIS. Beginningfrom understanding the importance of education and education policies, foremostprime objectives of educational policies are identified (studying existingliterature). Thereafter, we look for quantifiable variables to judge theprogress made in achieving the above objectives. Furthermore, variables thataffect the above variables are proposed. Regression is carried out to study thecorrelation and causal relation between the dependent and independent variable. Introduction: Humanshave always been knowledge seekers; exploring, learning, evolving and changingthe system suited to their comfort. The knowledge gathered over the years hasnot only helped humanity to prosper by utilizing available resourceseffectively but knowledge has also been a base to build on further knowledgeand bring about progress. One manifestation of the knowledge system isEducation, formally defined as process or means of learning and acquiringskills.
Inthe modern setup education becomes even more crucial both because of theintrinsic value it possess and for the marketable capital it equips anindividual with. The link between education and development can be betterunderstood if we segregate its utility at individual, social and nationalscale. For an individual, education is a mean of accumulatingmarketing human capital, which expands individual’s employment opportunity.Education also builds scientific temperament which is a way out of ignoranceand superstition.
Education ensures basic literacy skills of reading, writingand arithmetic to individual who is crucial both for human dignity anddevelopment. Education is an important component of HDI. In the modern worldeducation can be individuals USP to overcome all social and economic barriers. Atthe level of society, educated citizens have increased political participationand high social values which is important for workability of good ‘socialinfrastructure’ and building up of ‘social capital. Education also ensuressocial co-operation resulting in improved gender parity and high femaleworkforce participation.
Atnational scale education is an important component of factors determininggrowth.Educationis crucial for Innovation, R. Innovation and creation of ideas necessaryfor long term growth. It is one of the determinants of technological progress.(Solow Model)Neo-classicalgrowth theories stress on human capital accumulation through education andtraining. It is essential for sustaining growthEducationis also essential for tapping the demographic dividend through building upskills. Historicaloverview FRAMEWORKFOR ANALYZING EDUCATIONAL POLICY Evidentfrom the multidimensional nature of educational policy, it is clear that thereis no single clear indicator or determinant. An analogy can be drawn with theunderstanding of development.
Economist have more or less a good understandingof what is development and what isn’t; but when it comes down to laying downeither some precise parameters to measure the extent of development, no singleparameter is sufficient and same goes for its determinants. Similarly when itcomes down to education, we have a fair idea of its broad objectives, thecurrent condition and the required corrective measures but we don’t have acomprehensive analytical framework to analyze comprehensively educationalpolicy and achievements. Atthe core of the framework lies dividing educational policy into four basicpillars – Accessibility, Equity, Quality and Efficiency. The rationale behindchoosing them is as follows. Education, being one of the most fundamentalrequirements in modern society, any educational policy should foremost ensurethat education is within the reach of masses. Further access to it should befairly easy. Now, it makes little senseif the accessed education provides no return and neither fulfill any of itsobjectives.
Hence ensuring quality is again very crucial. Education,furthermore, is an important means of achieving capabilities that go a long wayin reducing inequality in several dimensions, caste, class, gender andregional. Education, although very crucial, shouldn’t involve any wastage ofresources and hence ensuring efficiency is also important.Itis believed that any educational policy objective can be classified and studiedunder the above pillars. Methodology:Ourmethodology is as follows-? VARIABLES TO JUDGE THE PARAMETERS? INDEPENDENT VARIABLES DETERMININGTHESE PARAMETERS? REGRESSION ANALYSIS TO CHECKHYPOTHESIS. Theframework has been build following a twofold process. Each pillar has beentaken considered one by one under the study.
Foremost, we come to the problemof how to measure the outcome/performance in all these objectives. We examinethe effectiveness of some of the existing measures and also suggest somemodifications/ alternatives. In the second stage we proceed to identify somemajor factors determining outcomes in each of the considered pillar. For thiswe carry out an extensive literature review extracting the various determinantsconsidered in the vast pool of research papers.
Through regression analysis wethen study how crucial are the understood determinants. In the process we alsoidentify some missing determinants. ACCESSIBILITY Oneof the most widely used measures for accessibility is Gross Enrollment Ratio.GER is defined as ratio of students enrolled in particular grade by the totalstudents eligible for that grade. The more is the number of students enrolled;it is believed to be an indicator of improved accessibility. Studyif the data shows that GER generally varies between 90 to100 in most of thestates.
The low variability in itself indicates that GER is not a very goodmeasure of accessibility. This claim is further strengthened by the fact thatit doesn’t take into account the student’s attendance. Several studies haverecorded high student’s absenteeism in primary schools despite high overallenrollment in the schools. Hence ameasure discounting GER for attendance would seem to be more appropriate. Thisanalysis anyway takes GER as a proxy for accessibility then through regressionanalysis of cross sectional state wise data of Indian states aims to determine thefactors that are crucial in determining improved accessibility.
Fig1 – Plot of GER and No of SCH per Block area (a) for all states (b) for 20selected states(TheGER data is from DISE flash statistics 2016) Aplot of GER primary students (class 1 to 8) on number of schools per block asshown in figure 1 (a) , shows a positive relation between the two, but therelation observed is not sharp and many data points are scattered away from thepositive cluster. Figure 1(b) plot the same relation for 20 selected states,again the relationship is positive but not very sharp. Oneof the reasons for observed scattering may arise out of the fact that thedivision of blocks is not uniform. Some of the blocks may in fact be very largein size compared to other blocks and hence similar number of schools may nottransform into improved accessibility. Toimprove the uniformity of parameter, no of schools per 10sqkm area is nowconsidered. Fig2 (a) – Regression of GER on SCH per 10 sqkm area; (b) – plot GER vs no ofschools per 10sqkm area for 20 states.
Thescatter-plot in fig 2(b) shows positive correlation between the GER and totalnumber of schools per block. This correlation became even stronger when thedata set was narrowed to 20 selective-states. To further ascertain thisrelation a regression is carried out with the dependent Variable as GrossEnrollment Rate (Total ENR/Total ELG) and the independent Variable: Number ofSchools per Area. The regression results clearly show that the model is jointlysignificant. Further the coefficient of dependent variable is also significantat 5% level of confidence. But the caveat here is the low R2 valueindicating perhaps some crucial variables are missing. Apart from the physicalpresence of school, the availability of transport infrastructure and the schoolinfrastructure in itself can be some of the factors that determineaccessibility. RegressionAnalysis carried out considering the Gross Enrollment Ratio and number ofschools in area gave a positive result which implies that the policies aimed atimproving the availability of schools implemented by the government so far,have been successful in bringing about improved accessibility accessibility ofeducation a possibility across the country.
EQUITYEducationalequity is a multidimensional concept. It spans from gender to caste/class toregional. For the case of building framework we just consider the case ofgender equality in education.
A measure for gender equity in education isGender Parity Index which is the ratio of female to male enrollment discountedfor the state wise gender ratio. Analysisshows that similar to GER, the variation in GPI data is even less, with most ofthe data points clustered between 0.9 and 1. One of the reasons for such littlevariation is the use of enrollment data itself, which is not discounted forattendance. Further enrollment data doesn’t capture all the dimensions ofgender equity. Say girls may be discriminated on the basis of resource beingmade available. The equity should further be reflected in the performance ofboys and girls in various areas, which again is not capture by GPI.
Thisanalysis anyway takes GPI as a proxy for gender parity and then throughregression analysis of cross sectional state wise data of Indian states aims todetermine the factors that are crucial in determining improved gender equity. Fig3(a) – Plot of GPI and proportion of schools with girl’s toilet. (b) – Regnresult for GPI on PRNSCHTT Thefirst factor that we consider for improved gender parity is the proportion ofschools with girl’s toilet. We carry out the regression with the state GPI asthe dependent variable and the state’s proportional of schools as theindependent variable. Regression model turns out to be significant, but thesign of coefficient opposite to expectation as shown in figure 3 (b). This isalso evident from the plot of GPI on proportion of schools with girl’s toiletwhich doesn’t yield any clear systematic picture.
The regression results showsthat some variable correlated with the independent variable has not beenincluded. Hence building girl’s toilet in itself may not result in improvedenrollment parity. Further as shown in several studies the quality of toiletsand their maintenance is also a major concern. Fig4 – (a) Plot GPI on State female literacy (all states) (b) for 20 states. The second factor that is considered is state femaleliteracy, which as an independent variable may act as a proxy for improvedgender consciousness translating into improved gender parity through moredirect participation and through indirect help.Regression of GPI on State female literacy turns out to beinsignificant, which is also evident from lack of any clear pattern in fig 4.
The regression results show that improved female literacy standalone will nottranslate into improved gender parity. QUALITY The paper also aims to understandfactors that influence quality of education. Fr this we use factors thatinfluence learning outcomes or more specifically performance on assessmenttests. The following two parameters were used for the same-? Abilityto divide a three-digit number by a two-digit number among children from agegroup 6 to 14 years? Abilityto read and comprehend a paragraph designed for students in the second grade. These constitute the dependentvariable for our regression analysis.Based on availability of datavarious factors were considered as independent variables. These includeeducation-level of instructor represented by proportion of teachers witheducation at the undergraduate level (or higher), number of teacher training institutes,infrastructure represented by proportion of schools with computers and/orplaygrounds and nutrition represented by proportion of schools with mid-daymeals. As highlighted in the PROBE report(2000), an environment that was conducive to and promoted learning was asimportant as the presence of an educational institution itself.
With this inmind we also regressed parameters for learning outcomes on overall literacylevel of the state. It is worth noticing thatavailability of ‘good’ infrastructure positively affects learning outcomes,this is indicated by the high value of the F-statistic and the low p-values forF-statistic and t- statistic. Hence, the above regressed models aresignificant.
Similar results as above can benoted from this regression, the high values of the F-statistic and the lowp-value for F-statistic indicate that the model is significant Further, the scatterplot shows apositive relationship between the proportion of teachers having undergraduate(or higher) education and the students’ ability to divide. Additionally,Regression of learning outcomes on proportion of schools with the mid-day mealscheme was found insignificant. This calls for the use of other indicators thatcan help in drawing a conclusive relationship between nutrition and assessmentof quality.The regression of learning outcomeson state-wise overall literacy levels was found to be significant, thusreinforcing the finding that environment has a role to play in the learningprocess.
The lack of availability of age-wisedata for learning outcomes and attendance records act as a limitation of thedata set used and consequently of the analysis itself. 1. EFFICIENCY Efficiency is defined as attainingthe maximum possible output from the used input or in other words producinggiven output with minimum input. In the following analysis totaldropout rates have been used as a dependent variable and State literacy rates have been usedas independent variables. The regression analysis carried out above suggests thatstates with higher literacy rates have lower dropout rates this conclusion fitswell with theory It can be inferred that children brought up in more educatedfamilies have greater motivation to study or greater support from their parentsfor education. Another interesting finding as can be noticed from theregression analysis is that with better management in schools have higherliteracy rates and those states have lower dropout rates. The aboveanalysis both reaffirms and deepens our understanding of the effect of variousfactors on efficiency.
CONCLUSION: The framework herein build provides a systematic approach todivide any educational policy objective foremost into one or more of thepillars- accessibility, equity , qualityand efficiency. The paper then highlights how regression based analysis can beused to study the impact of policy and also their linkages with other aspectsof education.The paper also reaffirms some of the earlier raised issuesof the inadequacy of the existing measures to ascertain educational outcomesunder various domains.The paper also highlights the need for a holistic approachtowards education and vouches for a comprehensive policy building focusing onevery linkages and aspect. REFERENCES: (TO BE ADDED)