There are two instances where row statistics are used, the first being for mean summaries. The other is for summary statistics, when row values have been put into ranges. When each table row has only one value specified for use in calculations, table statistics can be used. Table statistics are described later in this chapter.
Row Statistics and Mean Summaries
For an example of mean summaries, say you have ten questions and each asks respondents to choose an agreement rating for each statement from Strongly Agree (4) to Strongly Disagree (1). Additionally, you want to have a single table reporting the summary statistics for all ten questions.
You would choose Row Options for each of the ten questions and select the desired statistics for each. Table 13 in EXAMPLE.RPT is an example of such a table; see Example Files for more information. See Mean Summary Tables for a further description of mean summary tables.
You can use the # sign with an equal sign to signify which values should use a different value for row statistics. For example:
Q2_1 (1-5) #1=5,2=4,4=2,5=1
Q2_2 (1-5) #1=5,2=4,4=2,5=1
Q2_3 (1-5) #1=5,2=4,4=2,5=1
means, “use value 5 for 1, 4 for 2, 2 for 4 and 1 for 5 in variable Q2_1; use value 5 for 1, 4 for 2, 2 for 4 and 1 for 5 in variable Q2_2, and so on.” It is not necessary to use the pound sign for values that remain the same, such as 3=3.
| Important: The Count (C) feature should not be used for rows that are using the #. |
You can also use Glossary Variables to make “global” recodes and assign values so that the changes you make apply to a given variable whenever that variable is processed.
Row Statistics When Row Values Are Ranges
For an example of summary statistics when row values are ranges, consider the following question and its answer choice response ranges:
| HOW MANY TIMES HAVE YOU EATEN AT A FAST FOOD RESTAURANT IN THE PAST MONTH? | |
| Row Logic | |
| 0-5 | Q15 (0-5) |
| 6-10 | Q15 (6-10) |
| 11-15 | Q15 (11-15) |
| 16-20 | Q15 (16-20) |
| MORE THAN 20 | Q15 (20-99) |
If you want summary statistics at the bottom of your table for this question, row statistics must be used rather than table statistics. This is because table statistics are calculated using just one value per row.
You must add new rows for summary row statistics so that you can select that row and specify the statistic and variable name/data location for the calculation. To add new rows, use Add Row(s) in the Setup Tables dialog box. Once the statistics rows have been added, you can use Row Options to assign the desired statistics.
A line prints for each specified statistic. The first statistic prints on the line next to the row text. If there are multiple statistics, they are printed on additional lines beneath the first. You can write the row text in such a way as to have a label that lines up with each statistic. To do so, separate each statistic name with the pipe symbol (|) when you create the row text, so that each name prints on a separate line. For example, the row text:
Mean|Standard Deviation|Standard Error
prints so the row and its statistics appear as:
| Mean | 3.4 | 4.3 | 3.2 |
| Standard Deviation | .6 | .9 | 1.1 |
| Standard Error | .3 | .4 | .5 |
Using the above example, the row logic for the summary statistics might be:
Q15 (0-99)
As previously mentioned, you can also use Glossary Variables to make “global” recodes and assign values so that the changes you make apply to a given variable whenever that variable is processed.
Related topics:
Including logic for row/column statistics