NTC Literacy Survey 2008

Prepared for: North Texas Commission
By: Paul Ruggiere & Marielena Carpanzano,
Survey Research Center, University of North Texas

The following information contains only a portion of the entire NTC Literacy Survey. Click here to download the entire survey. Page 1 offers an overview of how the survey was conducted, while Page 2 displays the findings and results.

INTRODUCTION METHODOLOGY Sample Instrument Data Collection Analysis by Demographic Groups SAMPLE CHARACTERISTICS	FINDINGS Self-Assessed Reading Skills Difficulty Reading Various Materials Need for Help and Additional Training CONCLUSION
Download the entire survey

Page 1 of 2

I. Introduction

During the months of October and November 2007, the University of North Texas Survey Research Center conducted the Metroplex Survey. The Metroplex Survey is an annual survey of 1,200 residents of Collin, Dallas, Denton, and Tarrant Counties. Three hundred respondents are interviewed in each of the four counties. Organizations and researchers purchase time on the questionnaire to ask questions in their areas of interest. The Metroplex Survey offers a low-cost, annual method of exploring issues or for tracking responses over time.

The North Texas Commission (NTC) submitted questions designed to assess reading skills and needs for further reading education in north Texas. The questions covered the following topics:

Self-assessed reading skills and understanding of written English;
Experienced levels of difficulty performing common, every day reading tasks;
Needs for additional reading training.

This report details the findings of each question and provides further analysis at the end of the report that ties the information together.

II. Methodology

Sample

The conceptual population for the survey was all residents of Dallas, Tarrant, Collin and Denton counties who were 18 years of age or older and who reside in households with telephones. Random digit dialing (RDD) was used as the method of sample generation because it offers the best coverage of active telephone numbers, and it reduces sample bias. The RDD method ensures that:

the conceptual frame and sampling frame match;
unlisted telephone numbers will be included, and;
the sampling frame will be as current as possible, thus maximizing the probability that new residents will be included.

To ensure that only residents of the target counties were interviewed, telephone prefixes outside of the area were excluded. Screening questions were included in the survey instrument to exclude anyone outside of the target area.

The sample was stratified so that 300 interviews were conducted from each of the four counties. A total of 1,200 usable interviews were conducted and analyzed. In a purely random sample, 1,200 interviews yield a margin of error of 2.8 percent. This means, for example, that if 40 percent of the respondents answered "yes" to a question, we could be 95 percent confident that the actual proportion of residents in the population who would answer "yes" to the same question is 2.8 percentage points higher or lower than 40 percent (37.2 percent to 42.8 percent). Margins of error are not directly applicable to stratified samples but can be expressed to indicate the adequacy of the sample size.

The data were weighted by county population and race/ethnicity when presenting data for the full sample. Census estimates from 2006 were used to develop the weighting algorithm.

Instrument

The survey instrument was constructed using several sources. Metroplex agencies, companies and University of North Texas departments were offered the opportunity to place questions into the Metroplex survey. NTC staff submitted a set of questions for the 2007 survey designed to measure reading abilities. The questions administered for NTC are included in the Appendix.

Data Collection

Trained telephone interviewers who had previous experience in telephone surveys were used to conduct the survey. Each interviewer completed an inten¬sive general training session. The purposes of general training were to ensure that interviewers understood and practiced all of the basic skills needed to conduct interviews and that they were knowledgeable about standard interviewing conventions. The interviewers also attended a specific training session for the project. The project training session provided information on the background and goals of the study. Interviewers practiced administering the questionnaire to become familiar with the questions.

All interviewing was conducted from a centralized telephone bank in Denton, Texas. An experienced telephone supervisor was on duty at all times to supervise the administration of the sample, monitor for quality control, and handle any other problems. Data for the survey were collected from October 3 to November 11, 2007.

Analysis by Demographic Groups

Questions in the survey were cross tabulated with the demographic variables presented below. Variables included in the crosstabulations will be discussed in each section of the report.

Length of residence	Have children
Years of education	Income
Number of occupants in household	Own or rent home
Age of respondent	Ethnicity
Gender of respondent	Employment status

Whenever the responses to a single question are divided by demographic groups, the percentage distribution of responses within one group will rarely exactly match the percentage distribution of another group; there will often be some variation between groups.

The most important consideration in interpreting these differences is to determine if the differences in the sample are representative of differences between the same groups within the general population. This consideration can be fulfilled with a test of statistical significance. The Survey Research Center only reports those differences between groups that are found to be statistically significant.

III. Sample Characteristics

The characteristics of respondents in the sample are presented in Table 1. About one-third of the respondents had a high school diploma or less. Nine percent made less than $15,000 a year and 15.7 percent made less than $20,000 a year.

Page 2 >>