Statistics (People)

One of the most important tasks of a panel administrator is to check the composition of the panel regularly. The statistics features in the People → Statistics menu support you in this.

The People → Statistics → Panel overview menu provides an overview on the state of the panel population (“panel health”). The key data are presented on tabs.

Open

This tab contains a lightweight dynamic evaluation tool allowing you to split your panelist population on the fly by any master data or system master data you want.

• To make your choice, simply click on the Configuration button.

• You will also find a graphical representation of your results.

Please mind: Only activated panel statuses will be displayed.

Tracking data for the entire panel. For further information see chapter "Panel Tracking".

Panel surveys and master data surveys that are currently in the field.

• Clicking on a project title will take you to the field report for the respective project.

• The results are represented graphically, too.

Panel surveys and master data surveys that are not active anymore and that have exceeded their field period. The list is restricted to surveys with a status change during the last 4 weeks.

Development of the group size over time. It lists all the groups that you have explicitly selected to be monitored.

One of the purposes of group creation is to simplify the tracking of groups of panelists with characteristics that are important for the panel. The “Monitored groups” feature allows you to select such groups:

• These groups will be specially marked in the list of groups.

• Using the drop-down list “Monitored groups” in the extended search function of the list of groups, you can limit the display to the monitored groups or exclude them from the display.

• The group history shows the development of the goup size over time (Breadcrumb: People → Statistics → Panel overview → tab Group history).

To mark a group as described above, find the desired group in the People → Groups menu and click on the Edit group icon. This will open the dialog familiar from creating groups. Tick the checkbox “Mark group as 'monitored group'” and confirm by clicking on Save group.

The Details tab in the People → Statistics → Community statistics menu provides an overview on community activities.

The History tab shows the development of the various community activities over time. Using the options hidden beneath the Configuration button, you can perform drilldown analyses. Possible criteria:

• Panel group, website, language version.

• Optionally, the activities of deleted panelists can be included in the analysis.

• Type of activity: specific community contents (forum, blog, discussion, chat), section, type of community data (e.g. number of posts and average number per panelist), specific forum discussion (i.e. forum topic, if forums have been selected).

Open

The following tables provide explanations of the evaluation options and the activity data in the results display.

Evaluation criteria:

Evaluation results:

The Way of entry statistics allows you to split up the panel population according to way of entry, e.g. in order to check whether panelists with particular ways of entry are more or less active than average.

• You will find the function in the People → Statistics → Way of entry statistics menu.

The following tables provide an overview of the evaluation criteria and the available data.

Evaluation criteria:

Evaluation results:

On the Distribution tab, you can also create a graphical representation of your results. The data are selected in the Configurations dialog.

The Session statistics provide information on the activity of visitors and logged-in panelists on the external view of the panel. The following rules apply for session statistics and the underlying logging processes:

• The activities of all visitors are logged now, not only the sessions of logged-in panelists.

• A session is defined as the time span between the first and the last access to the website.

• If a person has logged in during a session (i.e. their data can be matched to a panelist account), then all page views of this session are counted as “performed by logged-in panelists”.

• In panels with multiple language versions, each session is assigned completely to the initial language version, i.e. the language version in which the session was started. This is only relevant, however, if there are multiple language versions and visitors or panelists frequently switch between versions.

Accordingly, the following evaluation options are available:

• The page views are split into views by panelists and views by persons who cannot be identified.

• The session data can be restricted to one specific language version.

• The session data can be grouped according to various criteria.

• Excel and CSV export is available.

If you wish to perform deeper analyses, you have the following options:

• The individual page views for different language versions of a session are available in the website log.

• The session data of individual panelists can be found on the Sessions tab in the detail view.

On the History tab, you can also create a graphical representation of your results. The data are selected in the Configurations dialog.

You will find the sessions statistics in the menu of the same name, People → Statistics → Session statistics, on the Details tab. The following tables provide detailed explanations of the evaluation options and the contents of the session statistics.

Evaluation criteria:

The People → Statistics → Panel statistics menu allows you to create frequency counts on various general criteria such as category or group membership, period of entry into the panel, panel status, master data and performance data.

The functional scope in brief:

• Split by two characteristics.

• Optionally, the results of several variables can be printed in one single table block.

• When using split variables, you can choose between various display modes: interlaced, non interlaced and cartesian product.

• Panel statistics for candidates status and language version status are predefined for several time intervals.

• You can download your evaluations in an Excel file.#

Please proceed as follows:

1. If the desired configuration has been created and stored for future use earlier, select it in the drop down list and trigger the evaluation

   process directly by clicking Load configuration. Panel statistics for candidates status and language version status are predefined for several time intervals.
   
   

2. Otherwise, start with defining the base.

3. Select the panel statuses which panelists should have to be considered for the evaluation.

4. In the “Display options” section, you can decide how the results should be displayed.
   
   

5. Select the performance data and/or master data which are to be evaluated. You can open and close the respective lists by clicking on the Arrow icons.

6. Start the frequency count by clicking on the Create statistics button.

The statistics is displayed. If you want to download it, click on the Excel export of statistics button. If you wish to repeat the evaluation in future, you can save the configuration.

You can use the following criteria to limit the base:

• Panel group category: If no specific category has been selected, the default category will be used.

• Panel group: If no specific group has been selected, the default panel group will be used.

• Language version

• Filter: The grouping filters are available.

• Split variables 1 and 2: Available are important system data (u_gender, pstatus, pinput, site_id, reg_code, pci) and all master data.

• Period of entry into panel

• Entered panel from / until

You can limit the evaluation to panelists with specific statuses. You have to select at least one panel status: If you do not tick any checkbox, status “active” will be used.

You can change the type of display:

• If multiple variables are used for evaluation, it is up to you to decide whether you want the results do be displayed in one single or in separate table blocks. If you want all results to be displayed in one table block, tick the checkbox “Display results for all variables in the same table block”.

• When using multiple split variables, you can choose from three display modes:

  • Nested: For each characteristic of the variable which is to be analyzed, the system will show all possible combinations of the characteristics of the first and second split variable in a table and count their frequencies.

  • Side-by-side: For each characteristic of the variable which is to be analyzed, the system will separately count the frequencies of the characteristics of the two split variables, their sums, and the percentage.

  • Cartesian product: For each combination of the characteristics of the variable to be analyzed and the split variable(s), the system will list frequency and the percentage individually.

There is a variety of important performance data at your disposal: u_gender, pstatus, pinput, md_status, md_num_updates, md_num_invites, reg_code, site_id, and pci.

Additionally, you can also use all tracking variables.

All master data with answer categories are at your disposal.

• For greater clarity, you can limit the list to specific master data categories.

• Optionally, it is possible for our support to configure the panel in such a way that text variables are available as well. But this is only recommendable for specific analysis scenarios.

You have the possibility to store a configuration you have made and to continue using it. The functions for entering the title and for selecting the evaluation period for the stored configuration can be found in the section “Saving the configuration of panel statistics” above the evaluation.

To delete an existing configuration, switch back to the settings dialog. In the “Configuration” drop-down list select the configuration to be deleted and confirm by clicking on the Delete configuration button.

In order to achieve a high level of quality, the collection of master data of all panelists has to be as complete as possible. Otherwise, problems will arise with creating groups and samples, and in particular with data evaluation. The ESOMAR guidelines therefore recommend that panel operators compile and publish a list of the master data for which values of all panelists are available. EFS provides daily updates of master data coverage figures:

• For each individual master data item, you can see for how many panelists a value is already available, and for how many this value is still missing.

• Alternatively, you can view the master data coverage for the master data categories of your panel.

• The coverage figures are updated once a day.

• You can decide for yourself whether to activate this function or not.

You will find the master data coverage of your panel in the People → Statistics → Master data coverage menu on the Master data coverage and Master data coverage (grouped) tabs.

Master data coverage for individual master data items

In the table header on the Master data coverage tab you will find the date and time of the most recent update. For every master data item, the table lists the same key information that is also listed in the master data overview: Variable type, title (clickable link to the edit dialog), internal variable name, data type, and master data category. In addition, there are three columns with the coverage values:

• Entry exists: Number of panelists for whom there is already an entry for the particular master data variable; i. e. either the panelists have already answered a corresponding question, or values have been imported.

• No entry exists: Number of panelists for whom there is no entry for the particular master data variable; i. e. the value of this variable is either “-77” (numerical variable) or “ ” (text variable). “0” is a valid value and counts as an answer.

• Percentage of existing entries: Percentage of panelists for whom there is an entry proportional to the sum of all existing panelists.

All panelists are taken into account, except for those with the status “deleted” and “tester”.

You have the following options for modifying or restricting the display:

• The titles of the table columns are clickable links. Clicking on any one of these reorganizes the table according to the entries in that particular column.

• In the Extended search, the available search and restriction criteria are internal variable name and title of master data, master data category, data and variable type, panel status, language version, and gender; i. e. you can have the master data coverage displayed exclusively for active panelists, for example, or for female panelists only.

• The View function allows you to display or hide columns and to change the number of lines per page.

The table on the Master data coverage (grouped) lists the coverage figures for each master data category:

• Entry exists: Sum total of the entries already available for master data of the respective category.

• No entry exists: Sum total of the missing entries for master data of the respective category.

• Percentage of existing entries: Percentage of existing entries in relation to the sum total of all required entries for master data of the respective category.

You can decide for yourself whether the master data coverage is to be assessed, and define a time for the daily update.

Please proceed as follows:

1. Switch to the People → Statistics → Master data coverage menu and open the Configuration tab.

2. If the update is already active, the times of the last and of the next scheduled update will be displayed in this dialog.
   
   

3. If the update has not yet been activated, tick the checkbox in the “Activate” field.

4. In the “Time of update” field, you can change the time of the daily update. The default time is set to 00:00 GMT. Please note that the setting will not be active before the next day.

5. Click on Save to confirm the setting.

In the People → Statistics → Panel quotas menu, multidimensional quotas can be defined for statistical evaluation: Target values for any combination of master data variables (e.g. age, gender, region) can be uploaded and compared with the actual values. From the percentages displayed for the individual cells you, as panel administrator, can easily tell whether it is necessary to recruit additional panelists for a particular segment or whether any cells are overpopulated.

The functional scope in brief:

• A panel can have any number of independent quotas.

• Up to 8 variables may be used in each quota.

• You can use all master data with defined characteristics, panel status (pstatus), language version (site_id), and gender (u_gender) as quota variables.

• The actual values of the quotas are regularly updated. For performance reasons updates are not performed in real time. However, you can trigger the update manually, if necessary.

• The quotas are currently used for statistics purposes only: No prospective panelist will be rejected at registration based on these quotas.

You will find the existing quotas of your panel in the People → Statistics → Panel quotas menu.

For each quota, the number of used variables and the date of the most recent calculation are displayed. The following functions are available in the “Actions” column:

• Edit quota: Opens the dialog where the properties of the quota are defined.

• Show cell values: Opens the overview of target and actual values.

• Upload target values: Opens the import dialog.

• Delete: The quota will be deleted after a confirmation prompt.

1. Click on the Create quota button.

2. Enter the desired title.

3. Select the grouping filter you wish to use for determining the actual allocation.

4. Select the desired quota variables.
   
   

5. Confirm by clicking on Add quota. You will be redirected to the overview of the quotas.

If you want to subsequently change quota properties, select the Edit quota icon in the quota overview. This will open the dialog shown above.

Create a CSV file for the import of the target values. This file should contain the following columns:

• a column for each quota variable selected in the previous step. The title in the first line is the name of the respective variable.

• a column titled “target”, containing the target values.

The first line contains the column titles. The following lines list the target values for all possible characteristics combinations of the quota variables.

1. Switch back to the overview of the quotas.

2. Locate the desired quota and click on the Upload target values icon.

3. Import the data as usual.

4. If the import was successful, a corresponding message will be output.

1. Switch back to the overview of the quotas.

2. Locate the desired quota and click on the Show quota allocation icon.

3. The target and the actual values will be displayed for each cell.

You can export the data by clicking one of the Export buttons above the quota allocation:

• Export (values and labels): This file contains the coded quota definitions along with the corresponding labels and values.

• Export (values): This file contains only the coded quota definitions and the values.