Summary tables will be automatically created for all analyses run with the exception of time series plots, histograms and reports that do not pertain to an individual constituent, such as Power Curves. A separate summary table is created for each type of test performed while running in either multiple-constituent or interactive mode.
Sanitas summary tables provide you with a detailed overview of a test series run for multiple constituents for any one type of test. For example, if you run a prediction limit analysis for five different constituents, the summary table will list a variety of prediction limit test result information (such as mean, standard deviation, limits, and exceedances) for each well and constituent. If you choose to rerun a particular test, Sanitas will detect that an entry already exists in the Summary Table for that particular well/constituent. By default, the existing entry will be replaced, but it is possible to add the new entry and keep the existing entry. Replacing the existing entry will produce summary tables containing only pertinent test results, thereby reducing the amount of manual selection/deselection of individual entries.
Several options are available within the Summary Table Setup window. You may select/deselect columns to be included/excluded in the table, set column widths which will be permanently stored for individual tests, sort the table by either constituent or well, or select the option to print only “significant” reports. More details are below.
A summary table entry is created for each test that is performed, whether the report is printed or not. If you want only a concise record of the test results and do not need individual reports, the tests can be run (in either multiple-constituent or interactive mode) without printing individual reports.
Once created, a summary table remains available through the Options/Summary Tables menu until another multiple-constituent run is performed for the same test in the same data file, or until the table is explicitly cleared (via a button in the Summary Table Setup Window).
Alternatively, you can choose to be prompted when a new analysis for the same type of test in the same data file is begun, to either overwrite the existing summary table or append the new results (see Prompt to Overwrite/Append Summary Tables on the Output tab of the Configure Sanitas window). If you clear or overwrite a table, the existing summary table's results cannot be reaccessed. Note: if you want to backup or otherwise access the stored summary table data outside of Sanitas, it resides in <sanitas>\util\<your data file name>\sumtab.mdb.
Summary table reports can be formatted, previewed, and printed by selecting the Options/ Summary Tables menu item and then choosing the type of test to be printed from the submenu (this submenu contains all summary tables that have been created for the selected data file). The Summary Table Setup window will then open, presenting a set of default settings that determine the content and appearance of the printout. By clicking the Print Preview button, you can view the report on screen, as it will be printed based on the selected settings.
Figure 5.21: Summary Table Setup Window
Changing any of the following settings will modify the format of the printed report:
Columns to Include: The summary table for each type of test will provide you with a specific variety of result data (see list below). Select the result data to be included on a summary table printout by checking the boxes to the left of the data names.
General Column Descriptions: Each Summary Table contains a number of columns unique to that particular test which may be found below. However, here is a list of some of the more general column headers that may be found in Summary Tables for most tests: Constituent (name of constituent), Well (well ID), Upper Limits (values are compared to an upper limit and determined out of compliance when there is an exceedance), Lower Limits (constructed for constituents such as pH that also test data for values which are too low), Date (sample date), Observ. (the reported concentration), Sig. (whether or not the test was statistically significant), Bg N (number of samples in background), Bg Mean (the average of the background data), Std. Dev. (standard deviation of background), %NDs (percentage of nondetect data in background), ND Adj. (an adjustment made to the background mean and standard deviation to account for data containing a certain percentage of nondetects), Transform (whether a data transformation was required to normalize data for a parametric test), Alpha (the significance level of the test for intrawell methods or the individual comparison in the case of interwell methods), Method (NP indicates nonparametric test, inter/intra indicates parametric test).
Summary Table-Specific Column Descriptions:
Box Plots: Std. Err. (represents differences across wells), Lower Q. (represents lower quartile of data), Upper Q. (represents upper quartile of data), Min. (minimum value), Max. (maximum value).
Sen’s Slope/Mann-Kendall: Slope (average rate of change in units per year), Mann-K (Mann Kendall calculated value), Critical (tabled value that the Mann Kendall calculated value is compared to).
Seasonality: K-W (Kruskal-Wallis test which determines seasonality), Chi-Sq. (Chi-Square test for normality), “df” (degrees of freedom).
Seasonal Kendall: Slope (average rate of change in units per year), “Z” (calculated value which is compared to critical values at different significance levels to determine whether trend is significant), 80% - 90% - 95% (various significant levels used for comparisons of absolute calculated value).
Control Chart: “h” (the concentration values are compared to this limit), “SCL” (the cumulative sums of data are compared to this limit), Deseas. (indicates whether data were deasonalized).
Rank Von Neumann: R(V) (Rank Von Neumann calculated statistic), Table (the tabled critical value).
Welch’s T-Test/Mann-Whitney: Calc. (calculated value based on data), 0.10 - 0.05 - 0.025 – 0.01 (various significance levels used for comparisons of calculated value).
Width (Column): Entering a different column width value in the box to the right of the test result name will change the width of any selected test result column. To prevent truncating data, the width of some columns may need to be manually increased to accommodate long constituent names, multiple background well ID’s, large numbers in the mean and/or standard deviation, etc. (This can be verified using the preview feature). Once the default settings are changed, the new settings will be permanently stored for a given test (for a specific data file).
Rows to Include: Which rows of data will be printed can be changed by scrolling through the list and clicking on a row to select or deselect it. A selected row is indicated by a checkmark to the left of the constituent name. Row selections are saved to the summary table to simplify future re-printing. For global selection/deselection right-click for a popup menu. An option to print only significant results in the Summary Table is described below.
Sorting: Click a column header to reorder the summary table information by selected/unselected, significant/not, well name or constituent name. Sorting on well names will use constituent as the secondary sort, and vice versa. This sort will also be reflected in the printed (or previewed) version.
a: This small button (only visible when applicable, meaning for summary tables that include an alpha column) provides access to the "Actual" SWFPR (technically analysis-wide false positive rate) report. Based on the values in the alpha column of the current summary table, and the retest plan in use, a total estimated false positive rate is computed. This value is distinct from the a priori site-wide false positive rate estimate discussed in conjunction with Power Curves in that it can incorporate a mix of parametric and non-parametric tests with varying sample sizes, and it is only available after the statistical analysis is complete.
Abbreviate Column Names: When selected, the printout column titles are printed in an abbreviated form. Using the abbreviated form decreases the total width of the columns (and consequently may allow for an increase in the font size used for the printout).
Portrait vs. Landscape: Controls the orientation of the report, which also affects font size and the number of rows per page.
Rows Per Page: By default this is a value that will comfortably fit on the selected orientation, at the default font size, but the number of rows is adjustable to accommodate other fonts etc.
Margin: This value determines the percentage of the width/height of the page that is devoted to a margin around the printed area.
Bold Significant Rows: When selected, the summary table will be printed with rows containing statistically significant exceedances or differences in bold lettering for rapid identification (non-significant rows will be printed in normal font style as always).
Significant Rows Only: When selected, the summary tables will print only those reports which were determined to be statistically significant.
Font: Any available system font and/or size can be selected. Not all printers have the capability to print all types; for more information on the capability of your printer, please refer to your printer’s manual.
Title: The title on the printed summary table can be altered by typing in the editable field containing the summary table name.
Stored Settings…: Displays the dialog box below, with options to reset all formatting settings (i.e. the options on the left-hand side of the window – not row selections) with factory defaults, or with the settings previously defined in a summary table belonging to another data file. This operation cannot be undone.
Clear Table…: Click to remove all contents from the currently displayed Summary Table (so that a new summary table can be created).
It is possible to edit the raw contents of a summary table using a database program, or to export it for further manipulation in a word processor or spreadsheet, for the purpose of integrating the results into a custom document or database, or changing the content or appearance of the table. The summary tables, in the form of a MS Access™ file, are found in <sanitas>\util\<your data file name>\sumtab.mdb. Opening this file requires MS Access™ or a compatible program, and the details will not be discussed here.
Warning: once a Sanitas summary table has been edited it may no longer be Sanitas-readable. We suggest making backup copies before editing summary table files.