Data, Reports and Papers

                  Numeracy Test                                                 Interview process                                               Literacy Test

Description of the instruments

• Household questionnaire: including modules on household composition, wealth, production and consumption.

• Woman questionnaire: including modules on family health and education, fertility choices, time use, household decision making, social participation, perceived confidence and self-efficacy. 

• Literacy and numeracy tests:  The literacy test includes reading the Hindi alphabet, Hindi syllables, Hindi words, non-sensible word construction in Hindi and paragraphs in Hindi and tests comprehension of the paragraphs. The numeracy test includes counting objects, single-digit recognition, double-digit recognition, identifying missing numbers in a mathematical series, addition and subtraction.

  • Verbal numeracy test: this test includes applied addition and subtraction question which are verbally asked and answered. 

  • Cognitive ability tests: These include the Forward Digit Span test, the Rapid Automated Naming with colors and the Wechsler Non-Verbal Aptitude test. 

  • Forward Digit Span (FDS): The subject is asked to recall the digits administered by the enumerator. In every consecutive attempt, the number of digits is increased by one.

  • Rapid Automated Naming (RAN) with colors: The subject is asked to identify primary colors in a random matrix arrangement. 

  • Wechsler Non-Verbal Aptitude: The subject is shown short sequences of geometric figures and asked to choose, among 4 proposed responses, the figure that would complete the sequence. 

              Respondent making a GARP choices                           Display of the GARP game                    Enumerators noting down choices                           

• Risk Aversion:   Through a series of low-stake incentivized binary choices between a sure amount and a gamble, we check for consistency in risk preferences and obtain a measure of risk aversion. In Phase II, this game is extended to include both a ordered set of choices as well as a set of choices out of order. 

• Preference Consistency: Through a series of  7 by 9 choices between boxes of biscuits, we test whether the subject choices are compatible with the Generalized Axiom of Revealed Preference, and if relevant, the degree of deviation (see series of pictures above)  

  • Knowledge and Confidence: Through a series of (incentivized) factual knowledge questions, we test whether the TA program increased knowledge of health and educational matters. In an extension of the game,  we allow the subject to "purchase" the answers of another subjects and using these alternative answers to revise their own answer. Using these data, we obtain a measure of "confidence" and are able to draw inferences regarding updating of beliefs.  

  • Household efficiency: We test for household efficiency through an investment game. We provide each member of the household with a randomly determined amount of cash and ask them to divide this amount between: (i) cash kept for themselves, and (ii) cash returned to a common (household) pool. The amount invested in the latter is multiplied by a factor larger than 1 but smaller than the number of participants. The decision is made in private. This game is repeated with different sets of household members.   

  • Household Bargaining: The bargaining game was played to measure the woman‟s bargaining power. In this game, the women and their spouses were asked to split Rs.30 between water tablets and rice independently and subsequently were asked to negotiate and arrive at a joint decision. The individual split would give us an idea of their individual preference for both goods. We then measure the bargaining power of the women from the variation between the joint decision and their independent decision.

Data

The data of this project can be downloaded via FIGSHARE

Research 

• Beyond Literacy: Affecting Women's Life through Adult Education (Annemie Maertens, Ashwini Deshpande, Christopher Ksoll and Sakshi Bhardwaj)

This policy brief presents a snapshot of the first rigorous evidence of the effectiveness of a Computer- based adult literacy program, TARA Akshar Plus. We find that TARA Akshar Plus is successful in raising the literacy and numeracy rates of its participants, and perhaps more importantly, impacts the lives of the students and their learners in a profound manner by increasing confidence and mobility and improving relations within and outside of the household.

• The Impact of a Computer Based Adult Literacy Program on Literacy and Numeracy: Evidence from India (Ashwini Deshpande, Alain Desrochers, Christopher Ksoll and Abu S. Shonchoy)

With over 700 million illiterate adults in the world, many governments have implemented adult literacy programs across the world, although typically with low rates of success partly because the quality of teaching is low. One solution may lie in the standardization of teaching provided by computer‐aided instruction. We present the first rigorous evidence of the effectiveness of a computer‐based adult literacy program. A randomized control trial study of TARA Akshar Plus, an Indian adult literacy program, was implemented in the state of Uttar Pradesh in India. We find large, significant impacts of this computer‐aided program on literacy and numeracy outcomes. We compare the improvement in learning to that of other traditional adult literacy programs and conclude that TARA Akshar Plus is effective in increasing literacy and numeracy for illiterate adult women. The working paper can be found here.  The published paper is available from here

• Changing One’s Mind? Updating of Beliefs among Rural Indian Women (Annemie Maertens and Christopher Ksoll)

How do people reassess their beliefs in the light of new information? Do people follow the Bayesian rule of utility maximisation? Or are there other aspects that people take into account when updating their beliefs? In order to better understand this process, experiments were conducted amongst rural women in Uttar Pradesh. The results and questions this study throws up are discussed here. 

• Adult Education, Knowledge and Confidence (Vinitha Varghese, Ashwini Deshpande, Christopher Ksoll and Annemie Maertens)

Illiteracy, in India and elsewhere, is largely a female phenomenon. The blog, published at Ideas for India, analyses a literacy programme aimed at adult women in India and finds that it has impacts beyond increasing literacy and numeracy. The general knowledge of participants improved and they were less likely to be over-confident about what they know – attributes that may contribute to better educating their children and absorbing new information. 

The results of this paper have now been integrated in a new working paper titled the Impact of Adult Education on the Next Generation

• Can Rationality be taught? Experimental Evidence from India (Christopher Ksoll and Annemie Maertens)

Standard utility theory builds on the concept of rational preferences, in particular a transitive and complete set of preferences. However, empirical studies have challenged these assumptions, establishing that individuals’ choices often do not conform with rationality. In this study, we use laboratory experiments in the field to first document the ability to make rational choices among (previously) illiterate women in rural India. We find that 55% of women make choices that violate the transitivity assumption in a series of choices over bundles of goods. When making choices between lotteries, 61% violate transitivity and 32% opt for dominated gambles. We then exploit the random assignment of the women to receive an adult education program to establish the causal effect of (adult) education on rationality. Reduced form impact regressions show that participation in the program decreases the probability of opting for a dominated option by 12 percentage points, or 32 percent, but there is no evidence that the literacy program affected whether preferences are consistent with transitivity. Finally, we estimate a structural model of decision-making in which the decision between lotteries is based on a noisy signal about the certainty equivalent from these choices. The results indicate that the literacy program significantly reduced the variance of the noisy signal for the newly literate women, suggesting a cognitive underpinning to the observed increase in rational decision-making.