There are three types of resources that can be browsed and/or analysed:
- Tabular datasets (Cubes)
- Survey datasets
- Other resources
Tabular datasets (cubes)
Cubes are typically multidimensional tables and consist of two or more variables (dimensions) and a 'measure'. Information is automatically displayed in a tabular format. Not all the available information may be displayed when the table is first opened, but by using the drop-down boxes above the table it is possible to explore the table further. Cubes are pre-defined by the data publisher and by nature contain aggregated information.
When a table has been selected from the browse list it will automatically be displayed in the right frame and 'Metadata' and 'Bookmarks' folders will be available in the browse list. The Metadata folder contains information about the actual table, which could include an abstract describing the table contents, or the name of the table creator, etc. The Bookmarks folder, if available, may contain a list of bookmarked views of the table that have been created and stored on the Nesstar server.
Item Level Metadata
When opening a table you may find a marker in the top left corner of some cells. This indicates that information is available about that cell. If you hover over the cell marker information about that cell will be displayed as a 'tool-tip'. There may also be markers for certain categories in either the the table header or row information. These indicate that information is available for those specific categories.
Tables can be examined by moving variables (dimensions) in or out of the table, and by creating a subset through the selection of certain categories. The 'Custom selection' option from the drop down boxes provides this functionality. Bar charts and time series graphs (if the table includes a 'time' dimension) are also available, and the table itself can be saved as a spreadsheet. The ability to display the information on a map may also be available.
For further information on viewing tabular (cube) data see the section: Manipulate tables.
Surveys represent a structured set of data in a variable by case data matrix.
A survey dataset, when opened, contains 'Metadata', 'Variable Description' and 'Bookmarks' folders. The Metadata folder contains information about the actual dataset, such as an abstract to describe the data, the name of the dataset creator, etc. The Variable Description folder contains a list of the variable groups (if created) and within these, information about each variable, e.g. question text, interviewers instructions. The Bookmarks folder contains a list of bookmarks that may have been created and stored on the Nesstar server.
The datasets in Nesstar are documented according to the Document Type Definition (DTD) for the DDI (Data Documentation Initiative), a standard for documenting survey and tabular (cube) datasets. The information is structured into major fields such as Document Description, Study Description (which may include title, abstract, keywords), Data Files Description (providing information about each data file), and Other Documentation. This information is available within the Metadata folder. Variable Description has its own folder. For further information about the DDI in Nesstar see also Metadata.
When you browse a dataset, metadata fields are displayed in the browse list. Click on a field name to see its contents displayed in the right frame. Links to downloadable User Guides may be found in the Other Documentation section within the Metadata folder. To view PDF files, you will need a PDF reader, for instance Adobe Reader or equivalent.
When you click on a variable within the Variable Description folder, information such as variable name, question text, categories and frequencies will be displayed in the right frame.
Exploring the data
Tabulation and Analysis (correlation and regression)
Survey datasets can be analysed by tabulating variables against each other and by performing correlation or regression analyses. Selecting the appropriate tab displays an empty table. The user then specifies which variables should be included in the analysis by choosing them from the browse list.
To view a graphical representation of the current tabulation or analysis, select the bar chart, time series icon, or if performing an analysis, the regression chart or scatterplot, from the top-right of the screen. When you click on the bar chart icon, several display options become available (e.g. stacked bar chart or pie chart) depending on the contents of your tabulation. To make a selection click on the graphical option of your choice.
Subsetting (selecting cases)
To perform an analysis on selected cases, for example, only male respondents, click on the Subsetting icon . Select a variable from the browse list on the left, and then choose 'Add to subset'. Highlight the category you want to include in the analysis from the right hand box and click on 'Add'. To include another category from the same, or a different, variable, click on the 'More' button. When you click on 'Ok', the analysis you have requested is automatically executed, and any future analysis will only be perfomed on this subset of cases. The message 'Filter is on' will also be visible.
To weight your data, or analysis, click on the 'Weight' icon . Predefined weighting variables are displayed in the box called 'Weighting variables defined in the dataset'. To select, click on your chosen weighting variable and move to the 'Weighting variables selected' box. The '>' and '<' signs move the variables from one box to the other. Variables that are not pre-defined can be used as weights by selecting them from the browse list. Click on 'Ok' to apply the weight(s). The analysis that you have just performed is automatically re-executed and weighted information will now be displayed. The message 'Weight is on' will also be visible.
Read more about weighting in Nesstar WebView.
Nesstar can also handle other resources, for example, Microsoft Word documents or PDF files. These files will appear alongside the other Nesstar datasets in the browse list, and typically include information about the resource (based on the e-GMS metadata standard) and a link (URL) to the actual document.