Variable - A measurable attribute of interest or data item; typically varies over time or between individuals or groups
Cell Contents - You must select one of these statistics to be displayed in each cell of your table.
Number of Responses - The number of people that provided a particular answer for the survey item.
Mean - An average. The mean is the sum of a set of values divided by the respondents/count of values (n).
Median - The value found at the exact middle of the set of all values in ranked order.
Missing Values - A value may be missing for a survey item if one or more of the following conditions is true (refer to your project codebook, data manager, or project director to determine what actually constitutes a missing value for your project):
- The respondent is never provided with the opportunity to answer the survey item (i.e., a skip pattern has occurred)
- The respondent does not answer the survey item (i.e., "Refused or Missing Data")
- The respondent indicates the survey item is not applicable to him/her (i.e., "N/A")
- The respondent indicates he/she does not know the answer (i.e., "Don't Know")
Output Options - These are additional statistics that you may choose to display in each of the cells of your table. (Note: Some of these items may not be available for your project.)
Cell Percentage - Percent of the total observations for a single cell. The circled percent shown below is a cell percentage. The circled percent indicates that black females make up 12.60% of the total number (2,651) of respondents. The upper left cell indicates this is a cell percentage. Also, the bottom right cell (total x total) will always equal 100% when a cell percentage is chosen.
Row Percentage - Percent of observations using the row total as the total number of observations. The circled percentage shown below is a row percent. The circled percentage indicates that 49.54% of the female respondents are white. The percentages in this row also indicate that 27.86% of the female respondents are black and 8.09% are multiracial. The upper left cell indicates this is a row percentage. The last column (total) will always equal 100% when a row percentage is chosen.
Column Percentage - Percent of observations using the column total as the total number of observations. The circled percent shown below is a column percent. The circled percent indicates that 49.01% of the white respondents are female. The percentages in this column also indicate that 50.66% of the white respondents are male. The upper left cell indicates this is a column percentage. The last row (total) will always equal 100% when a column percentage is chosen.
Standard Error - The standard error (S.E.) provides an estimation of the extent to which the statistic drawn from a sample differs from the population statistic. The smaller the standard error, the better the sample statistic is as an estimate of the population statistic. The standard error is a measure of sampling error; it refers to error in the estimate due to random fluctuation in the sample.
Chi-Square (PEARSON)- A statistical technique used to determine if a distribution of observed frequencies (counts) differs from what would be expected by chance. Many people also think of a chi-square test as a test of whether or not a row and column variable are related. The larger the chi-square statistic, the more likely it is that the row and column variables are related to one another.
Chi-Square (RS2 and RS3) - RS2 and RS3 chi-square tests are special types of chi-square tests that are used when the data are weighted. The purpose of the test and the way in which the chi-square statistic are interpreted are the same as with the PEARSON chi-square.
Confidence Interval (95% CI) - A confidence interval is also sometimes referred to as a margin of error. The confidence interval specifies the size of the interval that is expected to contain the true population value of the survey item of interest. Typically, a 95% confidence interval is used - meaning that the true value of the survey item will be contained in this interval 95 percent of the time.
Interquartile range - The interquartile range is a range between the lower and upper bounds of the middle 50 percent of the cases. To find the interquartile range, rank the cases and divide them into four equal groups. The two middle groups are the interquartiles. The interquartile range is the lowest and the highest value of these two middle groups.