Format Connect Results

The Format screen allows you to setup how you want your results displayed.  (see example)

You can have your results displayed in either an HTML table or an attractive, interactive graph.

If you have selected a large number of libraries and/or data elements, the table display will probably be more useful.

This option allows you to specify how to sort the results. You can sort by any of the data elements you selected in ascending or descending order. If you leave this option to ‘None’, then the results will be sorted alphabetically by library name.

If you want to sort by a particular data element but not display it on the graph, you should first select it from the Select Data Elements screen, specify that you want to sort by that data element on the Format Screen, and proceed to the graph. On the graph, you can click the magnifying glass and then select only the portion of the graph that displays the data element you want. You may also have to adjust the y-axis max values.

The Group By option controls how the results are organized into columns and rows, whether displayed in a table or on a graph. If you group by libraries, then libraries are will be placed in rows and data elements will be used as columns. If you group by data elements, the data elements are placed in rows and the libraries in the columns.

If you are using a graph, it is usually better to group by data elements. This is the default.  However, if you are using a table, grouping by libraries is usually better.

Web Connect can display 4 types of data. Only one type of data can be displayed at a time.

Regular data is what you normally expect to see. It is the actual values reported by the libraries.

Rank data shows what rank the regular value for that library is within that particular data element within the data set. Thus a library that reports the same circulation data to their state (then up to FSCS) and to PLDS, would rank differently on circulation in each of these data sets because the number of libraries in each set is different.

Percentile data is a broader way of expressing rank, but essentially expresses the same idea of position within a group. On large data sets like FSCS, percentile is often more meaningful than rank. For example, it is more meaningful to know that a library is in the 74 th percentile on collection expenditures than that it is ranked 2,238. However, on smaller data sets of less than 500 libraries and especially less than 100 libraries, rank is usually a better number..

Standardized:

Although standardized data is for stat gurus, it is still a very useful type of data. Standardized data tells how many standard deviations away from the average a particular value is.

A simple example will help clarify standardized data. 5 libraries spend an average of $100,000 on collections annually. One of them spends $75,000. Is that low? It depends on what the range of expenditures is among the 5 libraries. Compare these two scenarios:

Scenario 1 Scenario 2

Library A $150,000 $400,000

Library B $125,000 $5,000

Library C $100,000 $10,000

Library D $75,000 $75,000

Library E $50,000 $10,000

Both scenarios average $100,000. However, in Scenario 1 Library D spends the less than most of the other libraries. In Scenario 2, Library D is spending significantly more than the other libraries, but Library A’s huge expenditures skew the average significantly. The standardized data in each of these scenarios would look like this:

Scenario 1   Scenario 2

Library A 1.41   1.97

Library B -.71   -.62

Library C 0   -.59

Library D -.71   -.16

Library E -1.41   -.59

In Scenario 2, Library A is almost 2 standard deviations away from the mean. By definition of standard deviation (see illustration below), 2/3 of all libraries will fall within 1 (plus or minus) of the mean. 95% of all libraries will fall within 2 of the mean. Any library will a value outside of 2, is very exceptional.

Library A in Scenario 2 is still inside the red portion of the bell curve on the right, but just barely. In Scenario 1, Library C is right on the average line between the two yellow sections.

One of the best applications of standardized data is comparing two data types that normally are not comparable together. For example, you could look at Collection Expenditures and Collection Expenditures per Capita on the same chart to see if a particular library stands out on either of these measures and how far they stand out. Thus standardized data can help you quickly identify areas where a library stands out from others.

The other options that appear on the right side of the format screen allow you to customize certain aspects of the graph. They do not apply if you are working with a table. All of them can be changed on the graph itself and are provided here only for convenience.

Click the next button to proceed to the results.