1/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education Yovani Lubaale1, Goretti Nakabugo2, and Faridah Nassereka3 Abstract Academic performance in primary education plays a crucial role in obtaining further educational opportunities. Despite increased focus on addressing the inequality gaps in access to education, a number of studies have shown that children living in poor families with mothers who have low educational attainments experience less success, both in school and later as adults in the workforce, than children living in more advantaged circumstances. This paper analyses the effect of mothers’ education on the numeracy and literacy learning outcomes among children in Uganda. Mining data from the 2018 Uwezo Uganda Learning Assessment survey, we explore the influence of maternal education on learning outcomes. The findings showed that the proportion of children who demonstrated the ability of competently reading and comprehending a story of primary two level increased with increasing maternal education. Whereas only 13.6% of the primary four children whose mothers had never been to school were able to read and comprehend a story (the highest level in literacy assessment), more than four times (50.7%) of the children whose mother had above senior four qualification had similar abilities. A similar trend was seen with performance in numeracy where 31.9% of primary four children whose mothers had no education at all were able to attain the highest numeracy level, compared to 59.1% for children whose mothers’ level of education was beyond senior four. It was further observed that slightly more than one in three (35.6%) of the primary one/two children whose mothers had never been to school were completely non numerate compared to less than one in ten (9.0%) of the children whose mothers had studied beyond senior four who were non-numerate. Given the changes in access to schooling and impact on learning yielding from the global COVID 19 pandemic, whereas the data mined was collected before this pandemic, there is need for reflection on the home-schooling approach being proposed by government and other stakeholders considering that this is likely to benefit more children whose mothers have higher levels of education than those with less education or never. Keywords maternal education, learning outcomes, numeracy, literacy, data mining, Uganda Introduction Sustainable Development Goal four (SDG 4) is about provision of quality education. This goal ensures that all girls and boys complete free primary and secondary schooling by 2030. It also aims to provide equal access to affordable vocational training, to eliminate gender and wealth disparities, and achieve universal access to a quality higher education (UNDP, 2000). Equal access to education opportunities, the type or quality of schooling, resourcing/financing learners’ achievement in primary education plays a crucial role in obtaining further educational opportunities. Childhood education not only affects the achievement and happiness of a person at the individual level, but also shapes the labor force quality and capacity of innovation to determine the potentiality of the development of a nation (Zhonglu & Zeqi 2018). Academic outcomes in primary education plays a crucial role in obtaining further educational opportunities. A number of studies have shown that children living in poor families with mothers who have low educational attainments experience less success, both in school and later as adults in the workforce, than children https://doi.org/10.29173/iq1001 2/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 living in more advantaged circumstances (Hernandez & Napierala 2014, Gooding 2001). Maternal education is a key driver of education attainment of children (Birdsall, Levine & Ibrahim, 2005; Browne & Barrett, 1991 cited in Abuya et al 2018). It has been further revealed that the children of teen mothers who are able to go back to school and complete their education do show improvements compared to children of mothers who are unable to continue going to school (Dizon, 2014). This further shows that whatever level a mother goes back to school, her educational achievement has a direct positive impact on the learning outcomes of her children whether born to her before or after attaining higher educational levels (Dizon 2014). The effect of parental education has also been found crucial in learning a foreign language. Khodadady and Farnaz (2012) studied 1352 students’ performance in Iran and found that third graders whose fathers and mothers held secondary and higher education certificates performed significantly higher than those having parents with primary education in learning English as a foreign language. Against this backdrop, the team thought to mine the Uwezo assessment data exploring the effect of maternal education on the learning outcomes of learners. Objectives of the Paper The objective of this paper was to study learning outcome in literacy and numeracy in Uganda; Mining Uwezo assessment data to demonstrate the importance of maternal education to learning outcome. Research question Does the performance (literacy numeracy and ethno mathematics) of learners in primary one (P1) / primary two (P2) and learners in primary four (P4) differ by maternal education of the mother. Hypotheses to Test I. There is no difference in literacy, numeracy and ethno mathematics among P1/P2 learners by maternal education II. There is no difference in literacy, numeracy and ethno mathematics among P4 learners by maternal education. Methodology This section explains the source of data, the analytical framework and data analysis. Source of data The source of data for this paper is the 2018 Uwezo Uganda Learning Assessment (Uwezo 2019), which was the eighth learning assessment (literacy and numeracy) that was successfully undertaken in 324 sampled districts in October 2018. The survey conducted at household level was representative at both the national and district level and focused on assessing children aged 6-16 years. The mined sub-sample consisted of household data from 950 enumeration areas, involving 16,859 households with 45,676 Children of whom 29,430 were aged 6-16 years. Of these, 8,372 were in primary one and two (P1-P2) and 3,405 in primary four (P4) on which the analysis is based. The reason for selecting those in the primary one and two was because a lot of work has been emphasized at these low levels (Republic of Uganda, 2020). Selecting out the primary four children is because this is the class of transition from the lower primary thematic curriculum to a more subject based curriculum with a focus on use of English as the language of instruction. Based on the Ugandan education system, age six5 is the official age of entry into primary one and if the learner went through the system without disruption, then they would be in senior four at the age of https://doi.org/10.29173/iq1001 3/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 sixteen years (Republic of Uganda, 2008). However, due to other factors (like working mothers, household poverty, distance to the nearest primary school), children’s age of entry in school varies whereby some children begin Primary one as early as 5 years and others begin late at the age of 8 years while other repeat classes (RTI6 2018). About Uwezo assessment The Uwezo assessments aim at generating evidence on learning outcomes in basic literacy and numeracy that can be used to influence policy action, reforms and practices towards improving learning outcomes. The 2018 learning assessment adopted a two-stage cluster sampling design with households as the elements and EAs as clusters. Selection of 30 EAs in each of the 32 districts was done using probability proportional to size. Selection of households at the second stage was done using simple random sampling following which 20 households were selected from each of the 30 EAs per district. The survey involved assessing all children regularly residing in the household whether in or out of school using a simple primary 2 test in literacy and numeracy as well as collecting background information on the children and household. Data on selected SDGs was collected including that on the quality of drinking water, health and nutrition. Detailed explanation on the methodology can be obtained from Uwezo eighth learning assessment report (Uwezo 2019). The Uwezo assessment tests in literacy and numeracy are benchmarked against the Uganda Primary 2 curriculum. While developing Uwezo tests competencies in order of increasing difficulty are distinguished. For the literacy, tests assessed a child’s competence by determining their ability to; recognize alphabet letters, recognize and read commonly used words, read a short sentence and their ability to read a short story and comprehend it. The numeracy test assesses basic skills in terms of counting and matching, number recognition, operations with whole numbers and every day mathematics in form of word problems. During the assessment, each child is graded according to the highest level achieved. However, for both competencies, non-reader and non-numerate levels are established to categorize children who are unable to recognize letters of the alphabet or count and match and are regarded the lowest gradable level. The mined data categorizes the different learners by their level of performance in both literacy and numeracy. Data analysis The analysis compared learning outcome assessment among children in primary one and two combined and those in primary four with maternal education. At the bivariate level the final learning outcome of the child either in literacy or numeracy were cross-tabulated with the maternal education. To show the differences in the learning outcome, more emphasis for literacy was placed on those who attained the story reading and the non-literate while for the numeracy, emphasis was placed on non-numerate and those able to do division. An unadjusted binary logistic regression in which mothers is education was the independent variable while ethno mathematics and the three bonus questions were the dependent variables was run. The results are presented as odds ratios with the corresponding significance levels. The binary logistic (logit) regression model takes the form: ( )3,2,1,,education maternal 1 log setsetsetmaticsethnomathef p p i i =        − ……………1 https://doi.org/10.29173/iq1001 4/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 Where f=β0+ β1ethnomath+ β2set1+ β3set2+ β4set3+µ βi- the coefficients µ- The error term ρi- Probability of the child being vulnerable Analytical framework According to Zhonglu and Zeqi (2018), documented and undocumented daily experience shows that the impact of family socio-economic status on children’s academic achievement is not direct, but rather through the following two paths. First, families with relatively high socio-economic status will strive to secure quality educational opportunities for their children, such as those provided by key schools and markets in the system, which in turn will influence their academic achievements positively. Secondly, these key schools are the ones with excellent teachers. In turn, these learners not only have a direct impact on learning outcome but are also affected by the learning attitudes and behaviors through teachers and peers. The result is that better future learning outcomes and further educational opportunities. Thirdly, family socio-economic status affects children’s learning behavior and academic performance by affecting parents’ educational expectations towards children and their educational participation. Parents’ educational expectations and behavioral support for children are, to a certain extent, also affected by their socio-economic status, resources availability and parents’ ability. This analytical framework is a slight modification from Zhonglu and Zeqi (2018). Based on existing literature, this paper, has mined the Uwezo Uganda data to try and explain the learning outcomes of children based on their mothers’ level of education (maternal education). Whereas the formulation of this analytical framework used data from China, a country with different socio-economic and demographic characteristics to Uganda, it is very practical to the Ugandan situation. The analytical framework for this paper is displayed in Figure 1: The framework shows that the outcome for this analysis is learning competency in numeracy and literacy. There are direct and indirect factors that affect the learning outcomes of children. Those that affect the child directly include the child’s learning behavior, parents’ educational participation, and educational opportunities available. On the other hand, family social economic status is a background factor which can go through any of the three to affect the learning outcomes, for example, if the family belongs to the lower social economic status (SES), it may be hard in the era of online learning to afford a computer/tablet or smartphone. Lack of the any of these gadgets may impair a brilliant learner from achieving their potential. https://doi.org/10.29173/iq1001 5/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 Source: Modification of Zhonglu and Zeqi (2018) Figure 1: Analytical framework on the relationship learning outcomes Limitation of the Uwezo 2018 learning assessment data This analytical framework has not been implemented fully in this paper because not all the variables can be found in the dataset for example child learning behavior since Uwezo assessments do not collect data on child learning behavior or differences in educational opportunities. Little information is available on parental participation. Results of the study In order to understand the effect of education on the child’s learning competences, mothers’ level of education was categorized into six groups. Those who had never been to school at all, those who stopped in lower primary (primary one to primary four- P1-P4), those in upper primary (primary five to primary seven- P5-P7), those who never completed secondary, (senior to senior three- S1-S3), those who completed ordinary level (senior four- S4) and those who were educated beyond senior four and above S4. Table 1 shows the distribution of learners in P1-P2 and P4 by their mothers’ educational level. Table 1: Distribution of P1-P2 and P4 children by Maternal Education P1-P2 P4 Total Number % Number % Number % None 1791 21.4 656 19.3 2447 20.8 P1-P4 2232 26.7 974 28.6 3206 27.2 P5-P7 2991 35.7 1250 36.7 4241 36.0 S1-S3 732 8.7 299 8.8 1031 8.8 S4 422 5.0 158 4.6 580 4.9 Above S4 204 2.4 68 2.0 272 2.3 Total 8372 100.0 3405 100 11777 100.0 Source: 2018 Uwezo Uganda Learning Assessment Parents educational level attainment participation Family Social Economic Status Control variables: Gender, Language used at home, Region Differences in educational opportunities Child learning behaviour Child Learning outcome (Competence in Numeracy and Literacy https://doi.org/10.29173/iq1001 6/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 The distribution of children by maternal education shows that most children, had mothers who had acquired upper primary education. The table also indicates that one in every five children (20.8%), had mothers who had never been to school. On the other hand, slightly more than one in four, had mothers who had stopped in lower primary (27.2%). Less than 20% of the children had mothers who were educated beyond primary seven (S1-S3 8.8%, S4 4.9%, and above S4 2.3%). These results have a lot of implication on the performance of children. This needs deeper reflection more so in the light of the education sector response to the Covid-19 pandemic in which home schooling and learning materials are being distributed and parents are expected to assist their children learn at home (Republic of Uganda, MoES, 2020). Whereas some schools have been allowed to open, home schooling is going to take become another way of instruction even after Covid-19 comes to the end. The 2018 Uwezo assessment of learning outcomes in numeracy and literacy among children aged 6-16 years was benchmarked on primary two standard work. The various reading competences from this assessment are displayed in Table 2 and Figure 2a and 2b. Table 2: Literacy Outcome for P1-P2 and P4 children by Maternal Education Mothers level of education None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Highest level attained in English literacy Primary One and Two % % % % % % % Number Non-reader 52.5 50.3 43.6 27.9 24.1 13.3 44.2 3614 Letter 33.5 36.0 38.8 42.1 36.4 29.1 36.8 3013 Word 11.2 10.8 13.7 21.3 27.2 32.0 14.2 1163 Paragraph 1.9 1.5 2.3 4.2 5.3 13.3 2.6 211 Story 0.8 1.4 1.6 4.5 7.0 12.3 2.2 176 Primary Four Non-reader 8.3 7.6 7.5 4.1 3.9 3.0 7.1 237 Letter 35.6 33.3 27.7 20.9 20.1 9.0 29.5 984 Word 30.3 32.2 31.1 29.4 20.8 19.4 30.4 1014 Paragraph 12.2 12.3 13.1 16.9 16.9 17.9 13.3 444 Story 13.6 14.6 20.6 28.7 38.3 50.7 19.7 657 Source: 2018 Uwezo Uganda Learning Assessment https://doi.org/10.29173/iq1001 7/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 Figure 2a: Proportion of Non-readers by maternal education Figure 2b: Proportion of children who reached story level by maternal education Source: 2018 Uwezo Uganda Learning Assessment Figure 2a displays the proportion of children aged 6-16 years in P1-P2 combined, and those in P4 who were categorized as non-readers and could hardly read anything from the tasks given. A non-reader in this context implies a child with hardly any mastery of reading. Among the primary one and two children, the proportion of those who could not read anything was highest among children whose mothers had never acquired any education (52.5%) and lowest among children whose mothers had above senior four level of education (13.3%). The same trend is observed among the primary four learners, were about 8.3% of the primary four pupils whose mothers had never been to school were classified as non-readers while just 3% of the those whose mothers had been educated above senior four were classified as non-readers. The ability to read a story was the highest level that would be reached by a learner in relation to the literacy assessment. Results indicate that less than one percent of children in primary one and two whose mothers had never been to school were able to read a primary two story. However, this proportion grows exponentially as the mothers’ education increases. Slightly more than one in ten children in primary one and two were able to read a story whose mothers had above senior four level of education. Looking at primary four children overall 7.1% of children were categorized as non-readers. However, the category of children contributing most to this result are children whose mother had never attended school (8.3%), those whose mothers stopped in lower primary (7.6%), and those whose mothers were in upper primary (7.5%). It is surprising to note that only 13.6% of the children in primary 4 whose mothers had never been 52.5 50.3 43.6 27.9 24.1 13.3 44.2 0.0 10.0 20.0 30.0 40.0 50.0 60.0 None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Non-reader P1-P2 0.8 1.4 1.6 4.5 7.0 12.3 2.2 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Story reader P1-P2 8.3 7.6 7.5 4.1 3.9 3.0 7.1 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Non-reader P4 13.6 14.6 20.6 28.7 38.3 50.7 19.7 0.0 10.0 20.0 30.0 40.0 50.0 60.0 None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Story reader P4 https://doi.org/10.29173/iq1001 8/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 to school were able to read a primary two story. On the other hand, one in two of primary four pupils whose mothers had an education level of senior four and above were able to read a story. Results regarding numeracy are not different from those of literacy except that if one draws a trend line, the gradient for literacy will be stiffer than of numeracy. Among P1-P2 children whose mothers had never been to school, 35.6% did not reach the lowest numeracy level that required counting and matching numbers (0-10). This proportion declines gradually among children whose mothers had lower primary (30.5%), upper primary (26.1%) to the lowest proportion among those whose mothers had been educated beyond senior four (9%). Among P4 children whose mothers had never been to school, the trend is not smooth. However, there is a lot to note. None of the P4 children whose mothers had education above senior 4 was at the non- numerate level. Overall, about one in thirty among P4 children can be classified as non-numerate. The proportion of non-numerate children being high among mothers with upper primary education being higher than that of mothers without any level of education or lower primary is a surprise and requires further investigation. This can be by looking at data for others years to see if a similar trends or results will be noticeable. However, though the differences are observed, they are not statistically significant. Table 3: Numeracy Outcome for P1-P2 and P4 Children by Maternal Education Maternal education None P1-P4 P5-P7 S1-S3 S4 Above S4 Total P1-P2 % % % % % % % Number Highest level attained in numeracy Non-numerate 35.6 30.5 26.1 18.6 17.0 9.0 27.8 2264 Matching 30.7 33.5 32.5 30.6 29.1 24.4 31.8 2596 Number Rec 10-99 11.8 13.0 13.9 14.6 15.0 12.9 13.3 1086 Addition 7.5 9.0 10.9 13.0 12.9 12.4 10.0 816 Subtraction 7.9 7.2 8.9 10.1 11.7 17.4 8.7 710 Multiplication 2.2 2.0 2.7 3.5 3.6 10.9 2.7 224 Division 4.2 4.9 4.8 9.6 10.7 12.9 5.6 459 P4 Non-numerate 4.1 3.1 4.6 2.1 2.6 0.0 3.7 121 Matching 8.7 8.9 7.3 6.6 6.6 3.0 7.8 259 Number Rec 10-99 7.6 6.4 6.6 8.7 1.3 4.5 6.6 219 Addition 16.7 16.5 14.6 13.1 9.2 9.1 15.1 498 Subtraction 20.5 22.7 20.4 17.3 19.1 19.7 20.8 686 Multiplication 10.4 8.6 10.9 14.5 17.8 4.5 10.6 352 Division 31.9 33.9 35.6 37.7 43.4 59.1 35.4 1171 Source: 2018 Uwezo Uganda Learning Assessment https://doi.org/10.29173/iq1001 9/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 Figure 3a: Proportion of Non-numerate children by maternal education Figure 3b: Proportion of children who reached division level by maternal education Source: 2018 Uwezo Uganda Learning Assessment While reading a story and answering the comprehension questions is the highest level in literacy, division is the highest level of numeracy assessed by Uwezo. Among the primary one and two children, there was a gradual rise in the proportion of children who reached the division level. Among children whose mothers had never been to school, about one in twenty-five (4.2%) reached the level of division while among those whose mothers had been educated above senior four, slightly more than one in ten was able to reach this maximum level. This is similar to what was observed in relation to performance in literacy (12.3%) and numeracy (12.9%), that is, slightly more than one in ten of the primary one and two children were able to carry out division and story reading tasks. Among the primary four children, a reasonable number reached the division level irrespective of the mothers’ level of education. However, there are still noticeable differences. Whereas 31.9% of children whose mothers had never been to school were able to reach the division level, this raises to almost double (59.1%) among children whose mothers had acquired education above senior four. 35.6 30.5 26.1 18.6 17.0 9.0 27.8 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Non-numerate P1-P2 4.2 4.9 4.8 9.6 10.7 12.9 5.6 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Division P1-P2 4.1 3.1 4.6 2.1 2.6 0.0 3.7 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Non-Numerate P4 31.9 33.9 35.6 37.7 43.4 59.1 35.4 0.0 10.0 20.0 30.0 40.0 50.0 60.0 None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Division- P4 https://doi.org/10.29173/iq1001 10/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 Ethno mathematics and additional questions In addition to literacy and numeracy, there were four additional general knowledge /problem solving questions that are asked of every child being assessed. The Ethno mathematics question and a set of three additional general knowledge questions which needed logic and application of day to day experiences in problem solving including identifying and naming shapes were further administered to the children. The bivariate analysis between these four additional questions and maternal education is displayed in Table 4. Table 4: Performance outcome in Ethno Mathematics and General Knowledge Problem Questions by Maternal Education None P1-P4 P5-P7 S1-S3 S4 Above S4 Total Primary One and two Can do ethno mathematics 24.1 23.5 23.8 32.0 30.4 48.1 25.4 1982 Bonus question Set one 14.2 13.5 18.3 27.9 32.5 48.5 18.4 1495 Set two 7.7 8.5 10.4 13.6 14.9 27.4 10.3 822 Set three 6.1 6.1 8.2 9.2 9.7 16.7 7.5 604 Total 1791 2232 2991 732 422 204 8372 Primary Four Can do ethno mathematics 62.2 62.8 63.6 68.6 65.6 80.6 64.0 2069 Set one 46.8 45.8 49.2 59.7 58.9 77.9 49.7 1644 Set two 28.0 29.3 29.7 33.1 40.1 60.0 30.6 1004 Set three 18.2 18.2 21.6 27.9 25.3 37.9 21.0 689 Total 656 974 1250 299 158 68 3405 Source: 2018 Uwezo Uganda Learning Assessment In relation to ethno mathematics, the proportion of those who could do the ethno mathematics question did not follow a specific pattern. Nonetheless, there is some evidence to suggest that the higher the maternal education the higher the proportion of those who could do the ethno mathematics tasks. The results for the set one and set three question which was concerned with the identification of objects showed a similar pattern with ethno mathematics but at lower proportions. For example, 62.2% and 46.8% compared to 65.6% and 58.9% of the children passed ethno mathematics and set one question among mothers with no education and those above senior four level respectively. Performance on the two problem solving questions which mostly assessed logical reasoning reveal that maternal education has an influence on the children’s abilities. This revealed the largest percentage points gap of between 28% and 40% in this category of questions whose mothers had never been to school and children whose mothers educated beyond senior four. Binary logistic regression In order to explain further the differences in the performance, we carried out an unadjusted logistic regression with maternal education as the independent variable and Ethno mathematics and additional questions as the dependent variable. The results are displayed in Table 5. Children whose mothers had never been to school were taken as the reference category. The performance in ethno mathematics by P1-P2 children and that of P4 children reveals some interesting results different from those observed as the bivariate level. For the P1-P2 children, considering those whose mothers had never been to school, https://doi.org/10.29173/iq1001 11/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 significant differences in performance in ethno mathematics are observed after the mothers who completed primary school. The implication is that performance in ethno mathematics does not differ among children born to mothers who have never been to school and those in lower primary and upper primary though still results shows that even the little education received is important. Results are statistically significant when between children whose mothers have never been to school and those with above primary education. Among the primary four children, a significant difference was only observed among the children whose mothers have higher education that is above senior four (p=0.004). The meaning for this is that in primary four, performance in ethno mathematics does not vary much by the maternal education especially that this was primary two standard question. For the additional general knowledge questions revealed, children in primary one and two showed that they comprehended these questions better when mothers had some level of education than their counterparts whose mothers have never been to school. The odds ratios from the logistic regression show that the higher the maternal education the higher the ratios implying that as the maternal education increases, the better the performance even in the general knowledge /problem solving questions. On the other hand, little difference is observed in the learning outcome between a mother who has never been to school and those with lower primary education among the primary four children. In general, irrespective of the class of the child, little difference has been observed on these set of four questions, no differences were registered between children born to mothers who have never been to school and those who stopped in lower primary. Overall major differences in learning outcomes are observed among children when mothers are beyond primary. Table 5: Binary Logistic Regression of the General Knowledge Questions by Maternal Education Ethno Set1 Set2 Set3 Odds Ratio P>z Odds Ratio P>z Odds Ratio P>z Odds Ratio P>z P1-P2 None 1.00 1.00 1.00 1.00 P1-P4 0.97 0.652 0.94 0.504 1.11 0.379 1.01 0.945 P5-P7 0.98 0.811 1.35 0.000 1.40 0.002 1.38 0.008 S1-S3 1.48 0.000 2.34 0.000 1.89 0.000 1.57 0.007 S4 1.37 0.010 2.91 0.000 2.10 0.000 1.66 0.010 Above S4 2.92 0.000 5.69 0.000 4.51 0.000 3.10 0.000 _cons 0.32 0.000 0.17 0.000 0.08 0.000 0.06 0.000 P4 None 1.00 1.00 1.00 1.00 P1-P4 1.03 0.813 0.96 0.686 1.07 0.556 1.00 0.998 P5-P7 1.06 0.568 1.10 0.332 1.09 0.446 1.24 0.087 S1-S3 1.32 0.066 1.68 0.000 1.27 0.114 1.73 0.001 S4 1.16 0.446 1.63 0.008 1.73 0.004 1.52 0.050 Above S4 2.52 0.004 4.01 0.000 3.86 0.000 2.74 0.000 _cons 1.65 0.000 0.88 0.112 0.39 0.000 0.22 0.000 Source: 2018 Uwezo Uganda Learning Assessment https://doi.org/10.29173/iq1001 12/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 Discussion The research question to be answered by the study was “Does the performance (literacy numeracy and ethno mathematics) of learners in primary one (P1) / primary two (P2) and learners in primary four (P4) differ by maternal education of the mother”. Awan (2013) says that education is the most important factor which plays a leading role in human resource development. It promotes productive and informed populace and creates opportunities for the socially and economically deprived sections of society. Numerous studies such as Gooding, 2001; Rana; Nadeem; Saima; 2015) have revealed parental education more so maternal education to be a strong predictor of children’s education and behavior outcomes. With 22% of Ugandan women compared to 16% of men having no education (Uganda Bureau of Statistics- UBOS, 2017), it is worth investigating the effect of maternal education on learning outcomes of children. Mining the Uwezo Uganda data is one way in which this effect has been demonstrated in light of using evidenced based results. Mining the Uwezo Uganda data has been able to show that in terms of education outcome, children whose mothers have never been to school and those with lower primary are among most vulnerable. If the country is to get out of illiteracy and innumeracy to achieve SDG 4, there is need to come up with a special program to assist these children. There will be no development if these children are not assisted and as a country, it is the beginning of dualism. Similarly, according to Hernandez and Napierala (2014), policies and programs aimed at increasing educational and economic opportunities usually target the less advantaged especially those with low income. This analysis shows that not only poverty should be targeted but that is just part of the problem hence the need to look at other factors like parental education especially maternal education. Educational programs that aim improve education outcome by targeting the poor but not looking at the entire child will yield limited impact or take long to be realized without consideration of developing policies that promote women’s education. To focus simultaneously on both children and mothers will foster long-term learning and economic success for low- income families. Improving the maternal education should be the best way to improve learning outcomes of children. This will therefore bring in another aspect; how do we maintain children in the school especially the girl child in the era of universal primary education and universal secondary education. There is need to brake this cycle if Uganda is to develop. Whereas the study was conducted in 2018 just before COVID19 pandemic, there are a lot of lessons that can be learned in this new normal, especially in Uganda which has had long period of school closure. Since the beginning of Covid19 (March 2020), the government has been encouraging home schooling for learners due to the lock down of country. Considering the importance of maternal education and with over 80% of the children in Uganda with mothers who have not gone beyond primary seven, this causes a lot of concern. The lower levels of maternal education make it had for the parents especially the mothers to guide the children. Conclusion In mining the Uwezo 2018 data, we have been able to explore the effect of maternal education on child performance by looking at the literacy and numeracy outcomes of children. We have explored the three areas of assessment done by Uwezo that is literacy, numeracy and comprehension. The results from the mined data shows that as maternal education increases the likelihood for the child to reach a higher level of assessment increases. There were two hypotheses to be tested based in this paper. The first one was that there is no difference in literacy, numeracy and ethno mathematics among P1/P2 learners by maternal education. Results presented in Table 2, 3, and 4 and figures 2a, 3a and 3b showed that performance among the P1/P2 pupils varied by the maternal education. We therefore reject the null hypotheses and accept the alternative that https://doi.org/10.29173/iq1001 13/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 there is a difference in the P1/P2 pupils in literacy, numeracy and ethno mathematics by the maternal education. The second hypothesis was that there is no difference in literacy, numeracy and ethno mathematics among P4 learners by maternal education. Like the first hypothesis, results presented in Tables 2, 3, and 4 and Figures 2b, 3b, and 4b show that the performance among P4 learner varied by maternal education. Recommendation Data mining and utilization: The source of data for this paper was the 8th annual learning assessment, conducted in 2018. However, Uwezo Uganda has more than 8 datasets each having more than 40,000 learners. This is a big resource which researchers have not yet exploited to explain different reasons for declining levels of education in Uganda. Uwezo Uganda should try as much as possible to popularize this data to the general public including researchers, academicians and policy makers since it is easily available free from www.uwezouganda.org. This data can be mined for further analysis and will generate a lot of knowledge for the country. Additional analysis: More data mining to look at maternal education controlling for other factors like household socio-economic status, rural–urban residence and should be carried. Another study can mine the performance of children living with their biological mothers controlling for maternal education with those not living with their biological mothers. Government policy: Data was collected on child attendance of Early Childhood Education (ECE). This is in line with the government policy on promoting ECE. Uwezo data can be analyzed further and provide critical insights which can be used to inform policy. Similarly, further analysis of this data can be used to report on the SDGs. There is need for further analysis on learning outcome to inform policy and make meaningful decisions in this regard. This study has demonstrated the importance of maternal education, the government through either the Ministry of Education and Sports or Ministry of Gender Labour and Socio Welfare should resume adult studies that used to take place in the 1970s. However, these adult studies should target the use of technology to help parents, especially mothers. It can cover basics like creating and having an email, how to use online platform and so on. Limitation of the data: There were some important variables which could enrich analysis. We are proposing that one or two questions can be added to the assessment tool to enrich this analysis like mother /female caretaker occupation. This is because an educated working mother may not have enough time to be with her children like an educated mother who is not working. https://doi.org/10.29173/iq1001 http://www.uwezouganda.org/ 14/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 References Abuya B.A., Mumah J., Austrian K, Mutisya M., Kabiru C., (2018) Mothers’ education and girls’ achievement in Kibera: The link with self-efficacy. Accessed: 19th September, 2021, from https://doi.org/10.1177/2158244018765608 Awan, A.G. (2013) “Relationship between Environment and Sustainable Development: A Theoretical Approach to Environmental Problems” International Journal of Asian Social Science, Vol 3 (3) 741-761. Birdsall N., Levine R., & Ibrahim A. (2005) Towards universal primary education: Investments, incentives, and institutions. European Journal of Education, 2005, Wiley online Library. Accessed: 19th September, 2021, from https://doi.org/10.1111/j.1465-3435.2005.00230.x Dizon J (2014) Mother's Education Crucial to Academic Success of Children. Accessed: Jan 2021, from https://www.techtimes.com/articles/20024/20141113/mothers-education-crucial-to-academic- success-of-children.htm Gooding, Y. (2001) "The relationship between parental educational level and academic success of college freshmen " (2001). Retrospective, Theses and Dissertations. 429. Accessed: May 2020, from https://lib.dr.iastate.edu/rtd/429 Hernandez D.J. and Jeffrey S. Napierala, (2014), Mother’s Education and Children’s Outcomes: How Dual-Generation Programs Offer Increased Opportunities for America’s Families Khodadady E and Farnaz Farrokh Alaee. (2012). Parents Education and High School Achievement in English as a Foreign Language Rana M. Asad khan; Nadeem Iqbal; Saima Tasneem (2015). The influence of Parents Educational level on Secondary School Students Academic achievements in District Rajanpur Journal of Education and Practice www.iiste.org; ISSN 2222-1735 (Paper) ISSN 2222-288X (Online); Vol.6, No.16, 2015 Republic of Uganda, Ministry of education (2008) The Education (Pre-Primary, Primary and post-Primary) Act, 2008 ACTS SUPPLEMENT to The Uganda Gazette No. 44 Volume CI dated 29th August, 2008. Printed by UPPC, Entebbe, by Order of the Government. Accessed: October 29th, 2020, from https://www.parliament.go.ug/documents/1257/acts-2008 Republic of Uganda, Ministry of education and Sports (2020) Draft National Inclusive Education policy (Unpublished) Republic of Uganda Ministry of Education and Sports (2020): Framework for provision of Continued Learning during the Covid-19 Lockdown in Uganda RTI (2018) Uganda Early Years Study Milestone 3: Final Report. Accessed: October 29th, 2020, from https://assets.publishing.service.gov.uk Uganda Bureau of Statistics (2017) Women and Men in Uganda, Facts and Figures. Uganda Bureau of Statistics. UNDP (2000) Sustainable Development Goals, Goal 4, Quality Education. Accessed: January 13th, 2021, from https://www.ug.undp.org/content/uganda/en/home/sustainable-development- goals/goal-4-quality-education.html Uwezo (2016) Are our Children Learning? Uwezo Uganda Sixth Learning Assessment Report. Kampala: Twaweza East Africa Uwezo (2019) Are our Children Learning? Uwezo Uganda Eighth Learning Assessment Report. Kampala: Twaweza East Africa Uwezo Uganda (2020) Adapted Strategy (2020-23) approved by Uwezo Uganda Board of Directors on 20th March 2020 Zhonglu Li1 and Zeqi Qiu2 (2018) How does family background affect children’s educational achievement? Evidence from Contemporary China. https://doi.org/10.29173/iq1001 https://doi.org/10.1177/2158244018765608 https://doi.org/10.1111/j.1465-3435.2005.00230.x https://www.techtimes.com/articles/20024/20141113/mothers-education-crucial-to-academic-success-of-children.htm https://www.techtimes.com/articles/20024/20141113/mothers-education-crucial-to-academic-success-of-children.htm https://lib.dr.iastate.edu/rtd/429 file:///C:/Users/oschwart/Stokes/IQ/IQ45_3_4/www.iiste.org https://www.parliament.go.ug/documents/1257/acts-2008 https://assets.publishing.service.gov.uk/ https://www.ug.undp.org/content/uganda/en/home/sustainable-development-goals/goal-4-quality-education.html https://www.ug.undp.org/content/uganda/en/home/sustainable-development-goals/goal-4-quality-education.html 15/15 Lubaale, Yovani, Nakabugo, Goretti, and Nassereka, Faridah (2021), Learning outcome in literacy and numeracy in Uganda: Mining Uwezo assessment data to demonstrate the importance of maternal education, IASSIST Quarterly 45(3-4), pp. 1-15. DOI: https://doi.org/10.29173/iq1001 Endnotes 1Prof. Yovani M.Lubaale, Advisor (MEAL) and Data Management, Uwezo Uganda, Email: ylubaale@gmail.com 2 Dr.Goretti Nakabugo, Executive Director, Uwezo Uganda, Email: gnakabugo@uwezouganda.org 3 Ms.Faridah Nassereka, Senior Program Officer, Assessment, Action and Research, Uwezo Uganda, Email: fnassereka@uwezouganda.org 4 Some of the 32 districts have been subdivided since then like Arua, Jinja among others. 5 Section 10 of the 2008 Education act sub section 3a “primary education shall be universal and compulsory for pupils aged 6 (six) years and above which shall last seven years”. 6 RTI - Research Triangle Institute. https://doi.org/10.29173/iq1001 mailto:ylubaale@gmail.com mailto:gnakabugo@uwezouganda.org mailto:fnassereka@uwezouganda.org