Jump to ContentJump to Main Navigation
Literacy and MotheringHow Women’s Schooling Changes the Lives of the World's Children$

Robert A. LeVine, Sarah LeVine, Beatrice Schnell-Anzola, Meredith L. Rowe, and Emily Dexter

Print publication date: 2012

Print ISBN-13: 9780195309829

Published to Oxford Scholarship Online: May 2012

DOI: 10.1093/acprof:oso/9780195309829.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 26 February 2017

(p.157) Appendix A Literacy Assessment Methods

(p.157) Appendix A Literacy Assessment Methods

Source:
Literacy and Mothering
Publisher:
Oxford University Press

Background in Research at the Harvard Graduate School of Education

As discussed in Chapter 3, the literacy assessments that were used in our Four-Country Study are based on the theoretical framework and research methods developed by Chall (1983, 1996) and Snow (1983, 2010) and their colleagues at the Harvard Graduate School of Education (e.g., Chall, Jacobs, & Baldwin, 1990; Snow, Barnes, Chandler, Hemphill, & Goodman, 1991; Snow, Cancino, De Temple, & Schley, 1990).

The work of both Chall and Snow follows developmental and psycholinguistic models of literacy that view oral language skills as the basis for reading and writing development (for a review of the different models see Dickinson, Wolf & Stotsky 1992). In this theoretical model language and literacy abilities are assumed to vary along a continuum: One end corresponds to the casual, informal language that we use in everyday conversations, in e-mail exchanges, with familiar interlocutors, who share the same cultural and linguistic background knowledge; at the other extreme is academic (oral and written) language, which we use in more formal situations such as in schools and public settings (e.g., government offices, family-planning clinics, hospitals) and requires the ability to convey the same meaning to any interlocutor, regardless of context or background knowledge (Snow, 2010; Snow et al., 1990).

All spoken and written language tasks vary along this informal–formal/academic dimension. Obtaining an accurate measure of a person’s language and literacy abilities is therefore a complex endeavor that requires not one, but several kinds of assessments that allow researchers to examine these abilities separately (Chall et al., 1990; Snow et al., 1991).

(p.158) Chall’s and Snow’s studies with school-aged children reveal that academic literacy skills, whether oral or written, are learned primarily in formal school settings. Furthermore, these language skills are first acquired in the oral mode (Snow, 1990). In studies with bilingual (English–French) primary-school children Snow and her colleagues reported strong associations between tasks requiring different oral academic language abilities and school performance (Davidson, Kline, & Snow, 1986; Snow et al., 1990). As mentioned earlier (Chapter 3), an oral language task that Snow’s work has identified as a particularly strong predictor of school attainment is the ability to provide formal noun definitions. Indeed, results of Snow’s studies indicated that grade level and exposure to school English (her subjects’ second language) explained more variation in the bilingual children’s definitional skills than amount of exposure to home English.

Therefore, the ability to provide formal definitions, according to Snow and her colleagues (1990), is a specific skill that does not develop spontaneously with age, but needs to be taught and requires years of practice to develop fluency and consistency. Indeed, to achieve true mastery one needs to have knowledge of the definitional genre and of its various features as well as enough knowledge of the meaning of the specific words one is being asked to define (Snow et al., 1990, p. 93).

In contrast to Snow, who focused more on the relationship between oral academic skills and school achievement, Chall’s work focused primarily on reading and writing development. In other words, she was particularly interested in the academic language abilities required to comprehend and produce written text. Chall (1983, 1996) proposed that in learning to read every person goes through stage-like changes, each stage involving the acquisition of different subskills, all of which are necessary for becoming a full-fledged reader. In this model every stage involves changes in (a) the underlying processes, (b) the goals of the reading task, (c) the manner of reading, and (d) the materials that are read (Dickinson, Wolf, & Stotsky, 1992). Reading is therefore defined differently at every stage (Chall, 1983, 1996).

In the prereading stage subjects learn to recognize and discriminate letters and numbers and some elementary reading skills such as recognizing their names and other simple words. Once letter recognition becomes more efficient, the reader enters reading stage 1, where the emphasis is on decoding, both simple words and simple stories (around grades 1 and 2). By stage 2 (around grades 2 to 4), decoding skills have consolidated, so readers begin to focus most of their attention on the meaning of what they read. Reading stage 3 (around grades 4 to 8) is a crucial turning point in the reading process, according to Chall and her collaborators (1983, 1990, 1996). Readers progress from focusing on decoding to focusing on comprehending increasingly more complex and abstract materials such as science and social studies texts. Chall refers to this transition as progressing from “learning to read” to “reading to learn” (Chall et al., 1990, p. 11). She suggests that, if beginning readers do not make the transition to stage 3 successfully, their reading comprehension skills will develop no further and then rapidly start to decline. In fact, (p.159) for Chall, a possible explanation for the existence of so-called “functionally-illiterate” adults in the United States—adults with limited literacy skills despite having attended school—may be that they were unable to make this transition as young readers. Chall also cites evidence from studies that indicate that readers who are at risk for reading failure due to internal (i.e., learning disabilities) or external factors (e.g., home environments that do not support literacy development) also face particular difficulties during this major turning point. Chall called this pattern of deceleration that she observed in her studies and that of others the “fourth grade slump” (Chall et al., 1990, p. 15). When readers reach Chall’s stages 4 and 5, one can presume that all lower level skills (phonological and orthographic rules) have become completely automatized. All the attention is focused on developing complex comprehension skills such as drawing inferences and recognizing different points of view at stage 4 (around secondary school) and on synthesizing new information with one’s own at stage 5 (university level and beyond).

Between 1979 and 1990 Snow and Chall carried out a joint study that has great relevance for our research findings. The study consisted of an in-depth look at 30 low-income U.S. children—15 above-average and 15 below-average students—their families, and their teachers, starting when children were in grades 2, 4, and 6 and continuing, for 2 years, until they completed grades 3, 5, and 7 (i.e., longitudinal data within a cross-sectional design). The purpose of this study was to examine the effects of both the home and the school on low-income children’s language, reading, and writing development, immediately before and after the so-called “fourth grade slump” (Chall et al., 1990, p. 15).

Findings supported the existence of Chall’s “fourth grade slump.” But they revealed important differences between the two groups of readers. The timing of the slump in achievement and its intensity depended on the readers’ ability and on the reading skill that was tested (Chall et al., 1990, p. 33). Up until grade 3 high and low achievers performed at grade level, based on national norms. Low achievers began to decelerate earlier than high achievers, around grade 4, and this deceleration became more intense in grades 6 and 7. When examining achievement across the various reading skills, one finds that word definitions scores began to decrease earliest in both groups (after grade 3), whereas oral and silent reading comprehension scores declined last (Chall et al., 1990).

A follow-up study was conducted 5 years later, when subjects were in junior and senior high school (grades 7, 9, and 11) (reported in final chapter of Snow et al., 1991). The trends of the students’ literacy scores were similar to the ones identified in the elementary grades. In grades 7 and 9 students’ reading comprehension scores (oral reading and silent reading comprehension), though below norms, were still their strongest compared to the other reading skills (word definitions and word recognition) because they could still rely on context to aid their comprehension. By grade 11 the reading comprehension skills had also begun to slow down, particularly among the low-achieving readers. This pattern of deceleration in the higher grades does not mean that the low-income students lack the (p.160) ability to comprehend what they read, as it has been interpreted by the National Association of Educational Progress (NAEP) (Applebee, Langer, & Mullis, 1987). On the contrary, these patterns, according to Chall and her colleagues, suggest that if students have difficulties with word definitions, particularly those of academic vocabularies they need to learn beginning in fourth grade, and with recognizing and spelling these more complex and abstract words, it will eventually affect their reading comprehension as well (1990, p. 43). In their study Chall and Snow found that above-average readers were able to make the transition in fourth grade, but below-average readers could not. Below-average readers showed weaknesses in both word definitions and word recognition around the fourth grade, and thus their reading scores revealed an early and marked pattern of deceleration compared to national norms. Above-average readers remained strong in word recognition, spelling, and reading fluency. These skills, Chall and her colleagues (1990) hypothesize, seemed to help them compensate for their difficulties with word definitions. Their pattern of deceleration, therefore, began much later and was less intense.

In sum, longitudinal studies of low-income students done in the United States, comparing low-achieving and high-achieving subjects (Chall et al., 1990; Snow et al., 1991), add further support to the findings reported for the children studied in Morocco and Guatemala. Like Wagner (1993) and Gorman and Pollitt (1997), Chall and Snow found that retention of reading skills varies across time and across type of skill assessed. Some skills (reading comprehension) are retained longer than others (word definitions). But if students achieve, at least, a fourth-grade reading level making the transition from “learning to read” to “reading to learn” (Chall’s reading stages 3 through 5, grade 4 and onward)they not only retain but are able to improve their reading skills on their own, as long as they use them, even after leaving school.

Methods in the Field Studies

In the second part of this appendix we discuss the maternal interview and literacy assessment methods used in the Four-Country Study, indicating modifications specific to each study, and the interview and methods used in the UNICEF Nepal study of 2000. Table A-1 summarizes the assessment methods used in each country. It allows us to discern some of the similarities and differences across countries.

Methods in the Four-Country Study

Maternal Interview. In all four countries mothers were first interviewed by one of the researchers or one of the local research assistants. Interviews were always (p.161)

Table A.1 Summary of Data Collection Methods Used in Each Country

Country

Language Used for Testing

Maternal Interview

Reading Comprehension

(Scoring)

Noun Definition

Print Media Health Messages

Broadcast Media Health Messages

Illness

Narrative

Functional Literacy

Mexico

(rural)

Spanish

x

X

(0–4)

x

x

x

x*

-

Zambia

(urban)

Bemba and English

x

X

Bemba and

English

(0–3)

x

English

x

Bemba and English

x

Bemba and English

-

-

Venezuela

(urban)

Spanish

x

X

(0–5)

x

x

x

x

x

Nepal (urban and rural)

Nepali, Newari, and English

x

x

Nepali

(0–6)

x

Nepali

x

Nepali

x

Nepali

x

(Mother chose)

x (Mother chose)

UNICEF Nepal**

Nepali

x

x

-

Heath Knowledge

Health Behavior

-

x

* Mothers in Tilzapotla were asked to retell an illness they or their children had had as in the other countries. Only the first 2 minutes of each audio-recorded narrative were transcribed, to reduce time spent on transcribing and coding; this limited cross-national comparisons.

** See text for a more detailed description of assessments carried out in this study.

(p.162) conducted in the women’s homes in their native language or language of choice if they spoke more than one (as in Zambia and Nepal). The maternal interview covered the woman’s socioeconomic and educational background; the schooling of her parents, husband, and siblings; the current socioeconomic conditions in which she was raising her children; her reproductive and health behavior; her knowledge of child development; and her attitudes toward her own children.

The interview questions were not merely translated into the local language. In each country the interview went through a rigorous process of adapting the specific questions to the cultural norms and customs of the country under study, with the help of local researchers and assistants.

Literacy Assessments. The assessment of maternal literacy and language skills included reading comprehension, oral academic language proficiency (noun definitions), comprehension of health messages in both print and broadcast media, the ability to provide a health narrative in an interview situation resembling that used in clinics, self-report of reading practices, and functional literacy (everyday tasks for which women need to use their literacy skills).

Mothers’ responses were both audio-recorded and written down on coding sheets. The coding sheets were checked with the audio-recordings after the assessment to ensure the accuracy of the data that had been collected. Local researchers and research assistants were trained by Velasco and Schnell-Anzola in our various sites to ensure reliable data collection. A more detailed description of each literacy assessment task follows:

Reading Comprehension and Noun Definitions are fully described in Chapter 5.

Comprehension of Broadcast Media Health Messages. All sample mothers were played a tape recording of three or more health messages that were broadcasted regularly on the radio of their respective countries. In Mexico women were presented with four messages that informed them of the benefits of breastfeeding, and talking and playing with their young infant, as well as the importance of monitoring a young child’s weight gain. In Nepal the first message emphasized the importance of using oral rehydration salts (ORS, called Jeevan Jal in Nepal) when a person, especially a child, has diarrhea; the second was about family planning and mentioned Depo-Provera as an example of an effective contraceptive method; and the third explained how vaccinations can save children’s lives. In Venezuela mothers listened to five health messages (about AIDS, family planning, healthy eating, and alcoholism) to examine their oral comprehension. These health messages were also scored based on the number of idea units the mother could provide for each one (units predetermined by researchers prior to testing). For example, in this task, the maximum score for a Venezuelan woman was 68 idea units, with 14 and 5 idea units for each of the messages on AIDS, 16 for family planning, 19 for healthy eating, and 14 for alcoholism.

Comprehension of Print Media Health Messages. All mothers, except those with 5 or fewer years of schooling (illiterate or incomplete primary), were presented with radio messages in written form. Mothers were given print health messages, one by (p.163) one, and asked to take as much time as they needed to read it. Then they were asked to tell the examiner everything they could remember about each message they had read. As with the broadcast media health messages, maternal responses were coded for number of idea units they provided per message.

In Venezuela mothers read radio messages about diarrhea, AIDS, family planning, and cancer. Nepali mothers who had attended school (66 women in Patan and 41 women in Godavari) were presented with radio messages concerning the importance of vaccinating dogs against rabies, the importance of female literacy and the benefits of teaching children to defecate in a toilet (environment will look cleaner, less danger of spreading germs). The maximum total score mothers could obtain, for example, in Nepal was 27, with 8 idea units for rabies, 5 idea units for female literacy, and 14 units for defecating in the toilet.

Illness or Medical Narratives. This task was designed to simulate the response to questioning in a health clinic or hospital. Mothers were asked to recount two health crises they had had: one health crisis of the child that was part of our study, and one of the mother. If mothers were unable to recall a health problem of theirs, they were asked to narrate their first childbirth experience. Interviewers were instructed to prompt the mothers with general questions (e.g., “And then what happened?”). If a mother seemed to provide too short an account or was missing important information (what steps she took, what type of medication she used, if the person recovered, etc.), the interviewers asked more specific questions (e.g., “Did you take him to the doctor?”). A maximum of 10 specific questions per narrative were allowed.

Illness Narratives were collected in all countries except Zambia. Mexico (Tilzapotla) was the first country in which Illness Narratives were elicited from the sample mothers. Unfortunately, a decision to only transcribe the first few sentences (the first 2 minutes) of each audio-recorded narrative, to reduce time spent on transcribing and coding, limited these data for cross-national comparisons. In Nepal mothers produced rather short accounts of their health or that of their children. It is culturally not appropriate to speak too much about oneself. These were collected in Nepali and a subsample was translated into English to check the reliability of the coding scheme. In addition to coding the Illness Narratives for number of questions the examiner used to elicit them, they were also coded more holistically for their organization (whether or not a health practitioner would be able to follow and understand the narrative based on how the mother structured her account).

The most extensive coding of these narratives was done in Venezuela. Illness Narratives were coded holistically for three variables: organization (as in Nepal), control (how in control of the health crisis the mother seemed to be, i.e., did she get medical help quickly, did she follow the doctor’s advice), and the severity of the health problem (how close to death the child and the mother had been). In the case of the latter variable we asked a pediatrician to judge the severity of the children’s illnesses and a family doctor to judge the severity of the mother’s illness or (p.164) childbirth complication. Each feature varied along a 5-point scale: Higher numbers represented either a more organized narrative, a mother who was more in control, or a more serious illness. Two raters read and coded the narrative after reaching interrater agreement.

Functional Literacy. These tasks were designed to assess literacy skills usable in specific contexts of the women’s lives: (a) labels and signs (environmental print), (b) reading time, (c) oral arithmetic (functional numeracy), and (d) document literacy. These assessments were carried out in Venezuela and Nepal. Below we provide the Nepali assessment of functional literacy as an example because it was the most extensive one, given the high percentage of illiterate mothers in this country.

  1. a. Labels and signs: Mothers were shown seven different pictures depicting food labels (milk, salt) and street signs (danger, red cross, go slow, bus stop, family planning) they encountered on a daily basis. First they were asked to tell the interviewer what they thought the sign or label meant (identification). Then they were asked whether they had ever seen such a label or sign before and if they could remember where. (Locations such as “at the market,” “close to a school,” or “at the hospital”).

  2. b. Reading time: Mothers were shown pictures of clocks showing different times and asked to identify the time shown on each one.

  3. c. Oral arithmetic or functional numeracy: Mothers were read four arithmetic word problems and asked to come up with the correct answer. All word problems were adapted to situations in each country. For example, on a daily basis, Nepali women encounter such tasks as estimating the number of rotis she had to make for a family of two adults and two children, estimating bus fare, and estimating preparation time for a wedding. (Exact words are available on request.)

  4. d. Document literacy: These tasks were based on various documents that women in each country were familiar with, regardless of schooling. In Nepal, for example, women were presented with the following documents: (1) a Jeevan Jal (oral rehydration salts) packet, (2) a doctor’s prescription with words and drawings of the dosage the child has to take (commonly inscribed by Nepali doctors), (3) a child immunization card, and (4) a school registration form. In the case of the first task mothers were shown an enlarged version of the packet of ORS that is used in Nepal, which provides the instructions of how to prepare Jeevan Jal using both pictures and print. Then they were asked to read/look at the instructions and describe what they would do in the case of a 1-year-old child with severe diarrhea. Interviewers were asked not to provide any help to the mothers, not even the illiterate mothers. Responses were coded on the basis of nine idea units included in the five steps of preparation shown on the Jeevan Jal packet. These were as follows:

    • Step 1: Wash your hands with SOAP and WATER (two idea units).

    • Step 2: Pour SIX glasses of DRINKING water into a CLEAN vessel (three idea units).

    • (p.165)
    • Step 3: Pour the ENTIRE content of the packet into the water and STIR properly (two idea units).

    • Step 4: Give the child some of this water EVERY TIME he has diarrhea (one idea unit).

    • Step 5: Give the child mother’s milk, food and other liquids, MORE than he or she usually gets (one idea unit).

Mothers were also asked if they had ever used a Jeevan Jal packet, and if they had, whether someone had helped them interpret the instructions. Responses to these two questions were classified into one of three categories: (1) responses based on what the mother read on the packet we provided, (2) responses based on previous experience with a Jeevan Jal packet, or (3) responses based on both.

Assessment of Nepalese Children’s Literacy and Language Skills

One-hundred and sixty-four children, whose mothers participated in the study, were available for testing. They ranged in age 3.8 to 9.2 years of age, with an average of 6.7 years. Their grades in school included K-1, K-2, Class 1, and Class 2. The children´s literacy instrument was, in part, based on the SHELL battery devised by Snow, Tabors, Nicholson & Kurland (1995) for their studies of early literacy development.

Nepali and English Expressive Vocabulary Tasks. The children were shown a series of pictures depicting high-frequency (common) items (such as shoe and hen), and low-frequency (rare) items (such as hammer and rhinoceros) and asked to name them in whatever language they preferred. Although the words came from the Expressive One-Word Picture Vocabulary-R Test (1990), their high or low frequency in the Nepali language and culture was determined by a pilot test. Most children replied in Nepali and English, and appeared excited about displaying their English skills. The children were given separate scores for each language. No children preferred Newari, though most of the Patan children came from Newari-speaking homes. The scores on the high-frequency words ranged from 0 to 28 with an average of 21.1 for Nepali, and from 0 to 25 with an average of 14.6 for English. Scores on the low-frequency words ranged from 0 to 17 and averaged 8.1 for Nepali, and from 0 to 26 and averaged 4.9 for English. The high- and low-frequency scores for each language were summed to create a Nepali vocabulary score and an English vocabulary score. The Nepali score ranged from 6 to 41 with a mean of 29.3 (SD = 6.6); the English score ranged from 0 to 43 with a mean of 19.5 (SD = 9.6).

Noun Definition Task. This was the same task as was used with the mothers. Children’s responses were scored as to whether they included a superordinate term (i.e., qualified as a formal definition). The children’s scores were the number (p.166) of formal definitions that the child provided out of 10. Scores ranged from 0 to 10, but the average was only 2.4 (SD = 1.8), as compared with the average of 4.7 for the mothers.

Phonemic Awareness Task. The child was shown six sets of three pictures and asked to identify which pictures began with the same sounds (three sets) and ended with the same sounds (three sets). Their score was the total correct out of 6. Only four children could not identify any sounds, and some could identify all, but the average was 3.6 (SD = 1.55), showing that some children had not mastered this basic reading skill.

Sight Word Reading Task. The children were required to read aloud six words, presented out of context, that in Nepali were simple two-syllable words (such as “umbrella”). Scores ranged from 0 to 6 with an average of 4.3 (SD = 2.05) words. Half of the children (84) were able to read all six words.

Reading Task. Each child was required to read aloud and/or comprehend passages written by Schnell-Anzola (together with the most experienced Nepali research assistant) to match Nepali grade levels (K-1, K-2, and classes 1 through 3). At each grade level, children who gave correct answers to two out of four comprehension questions after reading the passage were considered proficient at that level and given the next-level text. Children who could not read the passages aloud were asked to listen to the passage and answer the comprehension questions. Children received a read-aloud score for the highest level passage they could read with fewer than three errors, and a comprehension score for the highest level at which they could comprehend a passage that they read or was read to them, answering two or more comprehension questions. Children who were not able to read at the most basic K-1 level were given a 0 for their read-aloud score.

The read-aloud scores ranged from grade 0 to 3 with an average of 1.4 (SD= 1.77). One half (80) of the children were unable to read the easiest text and received 0 as their read-aloud score.

Writing Task. The child was required to write his or her name and six simple dictated words (such as “cow” and “dress”). Scores ranged from 0 to 6 with an average of 2.6 (SD = 1.46).

The scores from the last tasks—phonemic awareness, sight word reading, dictated word writing, and read aloud—were composited into a single child literacy score using principal components analysis. The composite was the first principal component, which explained 70% of the variance and weighted the four variables similarly.

Methods in the UNICEF Nepal Literacy and Health Survey

A household survey, literacy test, and maternal health survey were administered orally in Nepali by native speakers to all 482 mothers of young children. (p.167) The household survey was used to gather demographic information and the results of that survey are presented in Appendix B. The literacy test and maternal health survey are described in more detail below.

Literacy Test. The literacy test measured functional, basic, and academic skills in reading, writing, and arithmetic. In each section of the test the earlier items measured more functional skills and the later items more academic skills. Measures were developed using guidelines from UNESCO, from Nepali functional literacy tests, and by choosing selected exercises from class 3 and class 5 textbooks for some academic skills.

Functional skills included picture, letter, and word recognition; writing names and addresses and letter dictation; number recognition; and time reading.

Basic skills included reading a sentence or an address written on an envelope, writing a dictated sentence or a description of a picture, and addition and subtraction.

Academic skills included reading comprehension where the woman read a passage from a class III Nepali textbook and then answered several questions based on the passage. Other academic skills measured include writing a short letter (two to four sentences) to a friend or relative, writing responses to questions based on text passages, three-figure addition and subtraction, geometry, fractions, and word problems from Nepali textbooks.

Scores for functional, basic, and academic skills were obtained on the reading, writing, and arithmetic sections separately and were summed to create a total score for the entire literacy test.

Maternal Health Interview. The maternal health survey contained detailed questions about women’s health-related knowledge and practices. The survey was designed to tap important public health issues for women and children in Nepal. Specific measures were adapted from prior health surveys including a study by S. Burchfield and colleagues at World Education, Inc. It also included measures contained in the Nepal Family Health Survey of 1996. Two composite variables, Health Knowledge and Health Behavior, were created.

Health Knowledge refers to women’s knowledge of vaccines, contraceptives, uses of medicines, and causes and preventions of HIV/AIDS. It was created using principal components analysis to combine mothers’ responses on the following items: the number of causes (M = 1.2, SD = 1.1, range 0 to 5) and preventions (M = 1.3, SD = 1.3, range 0 to 6) of HIV/AIDS she could name, the number of medicines she knew the uses for (M = 0.6, SD = 1.0, range 0 to 5), knowledge of the polio vaccine (M = 0.7, SD = 0.4, range 0 to 1), knowledge of what vaccines need to be given to children and when (M = 2.2, SD = 2.4, range 0 to 8), and the number of types of contraceptives she could name (M = 3.6, SD = 1.8, range 0 to 8). All measures were significantly positively related to one another with correlations ranging from 0.17 to 0.85 and an overall standardized alpha level of 0.77. The first principal component was selected that weighted all variables positively with the most weight given to knowledge of causes and prevention of HIV/AIDS, followed (p.168) by contraceptive knowledge, vaccine knowledge, use of medicines, and knowledge of the polio vaccine. This principal component contained 2.85 units of variance, 48% of the original variance in the included variables. The composite scores ranged from –2.89 to 6.90 with a mean of zero (SD = 1.69).

Health Behavior represents maternal behavior related to her and her child’s health during pregnancy as well as maternal behavior related to sanitation and keeping medicines in the home. It is the sum of a mother’s responses on the following nine items: whether or not she had prenatal care during her last pregnancy (N = 94), took iron tablets during pregnancy (N = 250), had all her tetanus toxoid vaccines (N = 130), delivered her last child in a health facility (N = 330), defecates in a latrine at the home (N =123), washes her hands after defecation (N = 156), uses iodized salt (N = 194), keeps any medicine in the home (N = 318), and treats unsafe water by boiling and/or filtering (N = 346). All measures were significantly positively related to one another with the exception of the relationship between taking iron tablets while pregnant and treating unsafe water. Significant correlations ranged from 0.10 to 0.39, with an overall alpha level of 0.70. Scores on the composite measure range from 0 to 9 with a mean of 4.96 (SD = 2.23). The sample size is 480 due to missing data for two mothers. Maternal scores on the Health Knowledge and Health Behavior composites were positively related (r = .56, p 〈.001).

Media Experience. As part of the health survey, mothers were asked about the frequency with which they read magazines or newspapers, listened to a radio, and watched television. These questions were considered important, as women could pick up health-related information from these sources. All three variables were related to one another and thus a composite was formed called Media. Seventeen percent of mothers said they read magazines or newspapers daily, 59% of mothers reported listening to the radio every day, and 48% reported watching television at least once a week. Mothers in Kaski and Chitwan were equally likely to read daily, yet mothers in Kaski were more likely to listen to the radio daily (t = 8.02, p 〈.001), and mothers in Chitwan were more likely to watch television weekly (t = –3.02, p 〈.01). A composite measure was formed in which 10% of mothers reported reading, listening to the radio, and watching television; 27% of mothers used two of these three forms of media; 39% used one form; and 24% reported no media activities. On average, mothers reported 1.2 forms of media exposure. Media exposure was positively related to maternal schooling (r = .56, p 〈.001) and maternal literacy skills (r = .46, p 〈.001).