Lighthouse is divided into four sections or panels. The table of contents appears in the left panel. You can navigate through the table of contents and select the indicators that you would like to include in your analyses. The right side of the window contains the Output Control. The Output Control allows you to view and alter the structure of your customized tables as you design them. The panel directly below the Output Control provides information on the dimensions available for the database.

In the Table of Contents panel, you can select the desired indicators to include in your analysis. The indicators are grouped into categories. To view the contents of a particular category or subcategory, click on the plus sign to the right of the category name. Continue clicking on the plus signs to search for the variables you would like included in your tables.

Click on the check boxes to select desired variables. Once you select a variable, the lower right panel will display the dimensions on which that variable was reported. You may also notice that other variables in the Table of Contents become disabled as you select variables. These disabled variables are incompatible with your current selection. You can change your selection by clicking on the checkbox to deselect an item. Should you wish to clear your entire selection, click on the Clear Selected button in the top panel. This will clear all selected dimensions and indicators. The Show Selected button will expand the table of contents to display all checked indicators. The Help button in the panel returns you to this help page.

Tips:

• The Table of Contents can be used to determine if a particular question was asked for a set of dimensions. If you would like to know when and where a question was asked, click on the indicator and then look at the lower right panel in order to determine the dimensions where data exists. If a dimension has an X next to it, the question was not a part of the assessment for that dimension.
  
• The Table of Contents allows users to explore how an assessment changed over time and across dimensions. When browsing the Table of Contents, you may notice that certain indicators are listed more than once. This means that the question differed slightly across dimensions. In most cases, the difference is due to the fact that respondents had different answer choices.

Back to top

The right panel builds the format of your data table as you select indicators. This panel presents different options of viewing the data and allows you to control the structure of your results.

Once you select an indicator from the Table of Contents, it will appear as part of the Output Control table in the right panel. Certain indicators have properties that determine if they belong in the table's columns or in the rows. Continuous variables can only be displayed in table columns. The indicators that fall under the scores topic are continuous and are therefore column variables by nature. In other words, a student's score is continuous because it could range anywhere along a continuum.

Topics other than scores generally contain discrete variables. A discrete variable has a select set of values. For example, when asked if a students has a computer in the home, there are only two possible responses- Yes or No. Discrete variables can be viewed in rows or in columns and are displayed as percents.

In the table, you will notice that the indicators have one or more button beside them. 

These customization buttons are used to modify the table structure in the Output Control.

Some indicators in the Output Control are marked with a flag in addition to the X. Indicators with a flag are discrete and can therefore be in rows or columns. Click on the flag to change a particular indicator from a row heading to a column heading or visa versa.

The collapse variables button is displayed for indicators with multiple discrete response options. This function allows you to customize response categories and is explained in more detail in the Collapsing Variables section.

Click on an X to remove a particular indicator from your table. A second way to remove an indicators is to deselect it from the Table of Contents. As you manipulate the table structure, you can change, delete, or add indicators and dimensions from the left and lower right panels.

If you would like to specify particular dimensions for your table, you may do so in the Dimensions panel. Once you are satisfied with your table design, click on the Make Tables button. Lighthouse allows you to span a maximum of 50 datasets in a single request. If your request will require data from more than 50 datasets, a pop up message will appear: 

 

If you did not specify particular dimensions and the indicators you selected are available for multiple dimensions, your request may generate several tables.  To limit the number of tables you wish to generate, click on OK in the dialog box and go to the lower left panel where you can select specific dimensions you wish to report on. Once you have narrowed your selections, click on the Make Tables button again to view your results. If your selection will result in more than 1 dataset but less than 50, a pop-up message will confirm your request and inform you that the web server might time out during the process. Click OK to process with the request or Cancel to refine your selection.

 

Note that the resulting tables may look slightly different from the output control. If the minimum sample reporting requirements for a particular row in the table are not met (there are too few records in that category), that row is omitted. If the minimum sample reporting requirement for an entire table is not met, you will receive a message indicating that there were no valid observations for the table specified.

Back to top

Collapsing Variables (Output Control)

The collapse variables option in the output control tables combines response categories for indicators with multiple discrete responses. For example, a user might want to analyze the relationship between a child's math performance and whether his or her parent(s) graduated college. The responses for parental education level include:

• Didn't finish high school
• Graduated high school
• Some post-high school education
• Graduated college
• Unknown

If the user wanted to focus specifically on college graduation, he may wish to group the responses into three categories:

• Graduated college
• Did not graduate college (includes Didn't finish high school, Graduated high school and Some post-high school education)
• Unknown

In order to collapse variables, click the collapse variables button to view the current response categories. Specify a name for the new collapsed category (i.e. Did not graduate college) and check the responses that you want included in this category (Didn't finish high school, Graduated high school and Some post-high school education). Then click Collapse. Your revised categories will be displayed. If you wish to create a second collapsed category, check additional response options and click Collapse again. If you wish to expand a category, check that category and click the Expand button. When you are satisfied with your categories, click Done to return to the Output Control view. The Output control now reflects the collapsed categories.

Example of Sample Output Table

	Grade 4 Math Composite
Parental education level	MEANS
Graduated college
Did not graduate college
Unknown

Back to top

Selecting Dimensions- Lower Right Panel

Dimensions indicate when, where and to whom an assessment was given. For example years, grades, subjects and jurisdictions. One dimension may be designated a trend dimension. If you select multiple values for this dimension, you can see how the values changed over time. To select a particular dimension, click on the checkbox to the right of its name. Because assessments report on different dimensions, selecting a particular value for one dimension may narrow the available values for other dimensions from which you can choose. Dimensions' values that are inconsistent with your other selections will be disabled and indicated by a gray x. If you designated particular dimensions to report on, some indicators in the table of contents may not be available for selection if they do not correspond with your specified dimensions. At any time you can clear your selections by clicking on the Clear Selected button.

Assessments are given nationally and by state. If you would like to view data for each available state and from the national sample, select All from the list of jurisdictions. If you wish to view only national data, select National from the list of jurisdictions. 

Tips:

• In order to make comparisons across values of one dimension, select the values of that dimension you wish to compare. Select a single value for each of the other dimensions for comparison.
  
  

Back to top

The indicators represented in the columns are the dependant variables and those in the rows are independent variables. According to the sample table below, about 78.79% of those who said they have an encyclopedia in their home also get a newspaper regularly. If we also wanted to find out what percent of people with a newspaper also have an encyclopedia, we could return to the Lighthouse and reverse the indicators to make encyclopedia the dependant variable.

 

Click on the Run button in the T-test column to assess whether the means or percents of two groups are statistically different from one another. A double asterisk indicates a high significance (<.01). A single asterisk denotes moderate significance (.01<x<.05). NS stands for Not Significant (>.05).

Tips:

• In order to compare results from 2 separate queries, run the first request. Then, leaving the Query Results window open, return to the Lighthouse Table of Contents. Generate the second request and click on Make Tables. The Query Results window will display the second query. Use the back button on the browser to view the first query again. You can view previous queries using the back and forward buttons. As you generate additional queries, you will be able to click through the output requests. Note that any additional requests you make will be inserted immediately following the output currently in the the Lighthouse Query Results window and all subsequent pages in the results window will be deleted. If you wish to keep a record of all requests, only click the Make Tables button when you are viewing your most recent data request.
  
  
• Lighthouse allows you to save results of a query as HTML file for future analysis. The file can be saved on your computer or network drive. (However, you cannot "bookmark" query results in the browser.) Once you save the HTML file, you can view the data and run T-tests on it even if you are not connected to Lighthouse. Some browsers may not display the graphics but this will not affect the functionality.
  

Back to top

In addition to Table based output, Lighthouse can produce easy to interpret graphs for means and achievment levels. When graphics are available, a Make Graph button will be enabled next to the Make Table button.

Lighthouse uses built-in intellegence to create an attractive and informative graph. For trend data, a line graph is automatically selected. Otherwise, a bar graph is drawn. If only a few variables are selected in the Table of Contents, then output tables will be combined by treating Dimensions like independent variables.

Tips:

• If you would like to include any of the graphs in a Microsoft Office Document, click the Save button below the graph and save the EMF file to a location on your computer or network drive. You can then go to the Insert menu in Word, Excel, or Power Point and choose Picture then From File.... You should then choose the EMF file that you saved. This will insert a high quality picture into your document.

Back to top