Selecting Libraries for Comparison

Selecting the libraries to compare is the first step in using Web Connect. The large box in the middle of the screen shows the list of the libraries that you have already selected for comparison. Just to the right of it are three buttons (‘select by peers’, ‘select alphabetically’, and ‘select benchmarks’) that will take you to new pages that give you special tools for choosing certain libraries. Just below the selected libraries box are two buttons that allow you delete some or all of the libraries you have selected. Finally to the left of the selected libraries box is a pull-down box that gives you special functions for saving and managing lists of libraries that you have created.  (see example)

Selecting libraries by peers is one of the most powerful features of Web Connect. Using this feature you can quickly create a list of all libraries that meet criteria that you define. For example, using the FSCS data, you could tell Connect to only show those libraries in Ohio, Pennsylvania, and Iowa that are municipal libraries with a population less than 20,000 and with more than 2 branches. More on

Select by Alpha

If you already know the libraries you would like to compare, this option allows you to choose libraries by name.  

If you already know the names of the libraries you want to compare yourself to, this is where you want to be. The Select State pull-down box lists all the states that are available in the current database (the one you selected earlier in the Setup screen). When you click on a state, all of the libraries in that state are displayed in the Libraries box in the center of the screen.

To select a library simply click on it. To select more than one library, hold down the CTRL button on your keyboard while you click . After you have selected all the libraries you want, click the Accept Highlighted Libraries button. You will be returned to the Select Libraries screen and the libraries you selected will be listed in the Selected Libraries box.

If you want to select libraries from multiple states, you will need to go through the process above for each state that you want to pick from. If you know that you will frequently use these same libraries, save them in a list so that you can conveniently load the list instead of manually picking them each time you want to use them.

Note: If you can’t find the library you want in the list, there are two probable reasons. First, the library may report under one name and be commonly referred to as another. See if the library is listed under a different name. Second, the library did not contribute data to the database. For example, the PLDS survey is voluntary and not all libraries participate.

Saved lists are lists of libraries that you know you will want to reuse. If you frequently compare your library with the same libraries, you should create a list with those libraries in it.

The Load List option takes you to the Saved Lists page, which lists all the lists you have previously saved. You can select one, click Load and the libraries in that list will be added to the existing list of selected libraries.

This option allows you to save the list of libraries that are currently selected for use later on.

You will be prompted to give your list a name. We suggest the name be descriptive so that you can remember what libraries are in the list.

This option allows you to delete a list of libraries that you have previously created.

You can only select up to 15 libraries for graphical comparison at a time. If you select more than 15, you will get a message on the results page letting you know that Connect cannot process your request.  You may select several hundred libraries for a table comparison, however.

Select Libraries by Peers

Selecting libraries by peers is a two-step process. The first step (Peer Finder Setup) allows you to choose broad criteria using both the statistical data elements and the categorical data elements. The second step (Peer Finder) allows you to fine-tune the list my adjusting your statistical data elements.

To select a data element as a criterion, click the plus sign of the tree control in the box on the left to expand the folder that contains the data element of interest. Next click in the box to the left of the data element’s icon. The box will become checked and the data element will appear in the box on the right, either under statistical data elements or categorical data elements depending on its time.  (see example)

Several other items will appear with the data element. For example if you choose Total Circulation as your variable, you would see the name of the library selected as the "Me Library" on the setup page, and the value for that variable in this case Total Circulation. You will see something like this:

Adrian Public Library: 73191

Minimum Value

Maximum Value

This information is provided simply as reference, and your criteria do not have to include the "Me Library’s" value. If you did not select a "Me Library" at the setup screen, instead of a Library Name you will see: "No me Library selected or no value available". This will also appear if you have selected a "Me Library", but there was no data for the variable that you selected.

If you selected a statistical data element, you will also see two input boxes side by side: the one on the left for the minimum value and the one of the right for the maximum value. Here are a couple of things to keep in mind:

1. Both boxes must contain data. If you don’t want to set a maximum value, then type in something huge like 10,000,000. Conversely if you don’t want a minimum value then type in 0.

2. You must choose at least one statistical data element, but you don’t have to choose any categorical data elements..

3. Remember the purpose of this feature is to narrow the query. If your range is too large, you could get back a list with thousands of libraries. Connect is not very fast when it has to pack around that many libraries. And in most cases, more than 20 libraries is often unwieldy and not very useful anyway.

When you’ve selected all the data elements you would like to use as criteria and entered the minimum and maximum values for each, you are ready to proceed to viewing the list and fine-tuning it. Click the next button to continue.

Peer Finder

The Peer Finder allows you to interact real time with the results of the criteria you setup in the step before. There are two parts. On the left is the list of the libraries based on the criteria you entered, sorted alphabetically. If the "Me" Library chosen at the setup screen is in the list of libraries it will display in red. You also will see a column of data for each of the statistical data elements you selected as criteria. If the entire title of the column is not visible, it can be expanded by clicking on its border and dragging it to the left or right. At the bottom, you will see the number of libraries your query generated. Remember, that for most purposes, more than 15 or 20 libraries are difficult to analyze and between 5 and 6 is usually optimal. Don’t feel like you have to have 300 libraries because its tough for Connect to meaningfully display them and its tough for you meaningfully analyze them.

On the right is a gray area with a minimum box (left side) and a maximum box (right side) for each of the statistical data elements you selected before. You can change the minimum or maximum value by typing directly in the minimum or maximum box and click the "Refine" icon. You will see the list of libraries shrink (or grow) as you make these changes.

You may expand your selection above or below the values you entered in the Peer Finder setup.  There is no need to go back to the Peer Finder Setup to change your range. If you wish to add or change variables, then, click the back button, add or delete a variable, and click next to return to the Peer Finder.

If no libraries appear in the Peer Finder, it usually means that no libraries met your criteria. Change your minimum and/or maximum value and click "Refine" or go back to the Peer Finder Setup screen and expand your criteria.

At this point in the software, many users think they’re through. There’s a "Graph It" button between the next and back buttons that will take the libraries and the data elements from the Peer Finder and drop them on a graph for you. However, in most cases the data elements just selected were used only for defining a small group of comparison libraries; they aren’t necessarily the data elements you wanted to know about each library. For example, you could have used Population and Income to define a group of libraries, but you’re interested at looking at staffing.  By clicking the next button, you will take the list of libraries created in Peer Finder and return you to the main Select Libraries screen. The list you just created with Peer Finder will appear in the Selected Libraries box.

In the Peer Finder ActiveX control, although you can highlight libraries, the Peer Finder uses all of the libraries in the list. It won’t select only the ones you’ve highlighted. You can delete selected libraries by using the Ctrl key.

Select Benchmarks

Benchmarks are statistical calculations based on groups of libraries that you define. An example of a benchmark could be the Average of the state of  Iowa. It could also be the average of the top circulating libraries. You can plot benchmarks on the graph just like you can real libraries. 

Although benchmarks can be graphed like real libraries, they are not actual libraries.  They are statistical calculations based on precise groups of libraries. For example a benchmark may consist of the average of all libraries in a given state. It may also consist of the average of the largest libraries, such as the Average of the 99 th Percentile for Population of the Legal Service Area.

When you choose benchmarks from the Select Benchmarks screen, you are not yet choosing the data elements for that benchmark. You are simply specifying how you want to identify the libraries to include in the benchmark. The actual data elements are selected on the data elements screen in a later step.

The first step is to choose a data element from the tree control on the left. If you choose a statistical data element, the box on the right will populate with all the possible benchmarks based on that data element. You will see 114 possibilities. 100 percentiles, 10 deciles, and 4 quartiles.

Percentiles are represented as Average of 88-89% for Population of the Legal Service Area with the lower bound (88) being the actual percentile. Deciles are represented as Average of 80-90% for Population of the Legal Service Area and are inclusive of the lower bound (libraries in the 80 th percentile) but not of the upper bound (libraries in the 90 th percentile). Quartiles are represented as Average of 75-100% for Population of the Legal Service Area. Like Deciles, Quartiles are inclusive of the lower bound but not the upper bound.

If you choose a categorical data element, you will see a benchmark for each value in the data set.

For example if you choose Legal Basis Code in the FSCS data set, you will see 9 benchmarks for each of the possibly legal bases codes found in the data.

You can select the mathematical method to calculate the benchmark—average, median, or sum.

The average method will average all the values for the data element that you select later on for all the libraries included in the benchmark. For example, you choose the state benchmarks of Average for New York, New Jersey, Ohio, and Florida, and in the select Data Elements screen later on, you choose % of Income from Local Government. The graph will average the % of Income from Local Government for all the libraries in each of the 4 states and display those results.

The median method will find the middle value for the libraries in the benchmark. Medians are helpful because they are not skewed by extreme values in the dataset.

The sum method will total all the values of the selected data elements for the libraries in the benchmark. In the case of a ratio data element, such as Collection Expenditures per Capita, the sum method will first sum the numerator and the denominator and then divide them to return the result.

To select a benchmark from the list of available benchmarks, simply click on it. To select more than one, hold down the CTRL button on your keyboard while you click. After selecting your benchmarks, click the Next button to return to the Select Libraries screen..

Note:

Unlike real libraries, which are actually stored in the database, benchmarks are calculated on the fly depending on what you select. It takes more time to look up the appropriate values for the libraries and then perform the benchmark calculation. Consequently, graphs and tables will take much longer to generate when you have included benchmarks, and the more you include, the longer it will take. In some cases it can take more than 45 seconds to pull up a graph with several benchmarks and several data elements. Also note that Average benchmarks are faster than Median benchmarks which are faster than Sum benchmarks.