The example table displays the results of the question, "How important is it to you to encourage people to use public buses and light rail within cities?" This question serves as the table's "focus variable." The question was truncated to "Encouraging public transportation" to serve as the table's title. Note that the variable label, ENV4, is included to the left of the title in order to allow the reader to cross reference with the full text of the question, which can be found in the section of this report titled "Survey Instrument." This is also the same variable label used in the raw data files on disk (in SPSS, MS Excel, or whatever format you requested). The number 27 under the variable label ENV4 indicates that this is the 27th banner table created for this survey.

The answers that respondents were offered in response to this question were "very important," "somewhat important," "not very important," and "not at all important." Note that these answer categories were truncated in the table headings in order to save space. In all survey questions respondents are allowed to volunteer "Don't Know" (DK), "No Response" (NR), or "Refused" (REF) as valid answers. The answer categories, along with "TOTAL"(for total number of respondents), serve as the column headings for the banner tables.

Under each of the column headings is a key for the rest of the table. Thinking of the entire table as a grid of cells, each cell contains 3 numbers. The furthest left, "Count," gives the raw number of respondents in each cell. The middle number, "Col," gives the percentage of respondents in a column in each cell. The third number, "Row," gives the percentage of respondents in each row. Columns read vertically down the page. Rows read across horizontally. The example interpretation provided on the next page clarifies the meanings of counts, column percentages, and row percentages.

The top row of numbers displays the "top line" results of the table's focus variable. Unlike the other cells in the table, the cells along the top row contain only two numbers; the "Count" and "Row" percentages. In this example 456 of 813 total respondents, or 56.1%, indicated that "Encouraging public transportation" was very important to them and 280, or 34.4%, said it was "somewhat important."

The variables displayed in bold along the left side of the table, referred to as "stubs," are questions that are cross-tabulated with the table's focus variable. In the example provided, the stub variable is "Sex." The totals for "Sex" are shown on the far right: 315 of the 813 respondents, or 38.7%, were male, and 498 of 813, or 61.3%, were female. Cross-tabulating "Sex" with the focus variable allows us to see differences and/or similarities in the response patterns of male versus female respondents. Note that sometimes the labels for stub variables and their categories are truncated in order to save space; their full text is available in the "Survey Instrument" section.

Example Interpretation: This example refers to those females who indicated that encouraging public transportation was "somewhat important" (the grayed cell). There are three interpretations of that cell:

1. Count: 168 of the 498 females who responded to the survey said that encouraging public transportation was "somewhat important."

2. Column: Of all 280 people who said that encouraging public transportation was somewhat important, 60% were female.

3. Row: Of all 498 female respondents, 33.7% said that encouraging public transportation was "somewhat important."

When interpreting percentages in words, statements should be phrased in terms of the denominator, that is, "of all those who said 'somewhat important'" for the column percentage in example 2, and "of all females" in the row percentages in example 3. Additionally, note that women did not differ substantially from men in their likelihood of saying that encouraging public transportation is somewhat important, since 35.6% of males, compared to 33.7% of females, gave such an answer.

Feel free to contact OSRL staff for assistance with interpreting the following tables.