After a project loads, you will find the dictionary for the loaded project under the project pane. In Harmoni, this is the Project Tree.
A project contains different items or variable types. All variables can be identified by their corresponding icon.
In this article
- Project tree filters
- Project tree Discover descriptors
- Project tree glossary-icons and colors
- Reset project tree
1. Project tree filters
Project Tree filter options allow you to quickly shrink the tree to the type of item required for either project design or analysis. Filters options are available at the top of the project tree.
- Filters reduce the items displayed.
- Filters are context-sensitive; options change depending on the type of items in your project.
- Applying filters shrinks the project tree; it only shows the selected type.
Grids – use this filter to show only grids.
Measures – use this filter to show only measures.
Verbatims – display verbatim comments only.
Weights – use this filter to show weight measures.
Projections - use this filter to display projection items.
Linked axes - use this filter to show linked axes.
Visible – use this filter to show only the visible items in your project (items flagged as hidden will not be shown).
Hidden – the reverse of the above. Shows only items you have flagged as hidden.
Common – where sources have been combined, this filter can be used to show only the common items that exist in all sources.
Uncommon – the reverse of the above. Where sources have been combined, this filter can be used to show only uncommon items.
Error "broken constructions" - Harmoni automatically identifies "broken" constructions as a result of a data source being removed or overwritten. Invalid constructions are shown in red in the project tree. You can also identify these by using the ERR icon under the Project tree filters.
New – use this filter to show new items. This filter will show any new variable or construction added to the project since the last time it was published. New items are shown in bold in the project tree. New items can be from a data source or new construction in the project tree. After publishing, new items will no longer be new.
Unsync-ed Items - use this filter to show constructions that have not been synced. The blue font also identifies items.
2. Discover Descriptors
Discover is a tool that lets you find the variables that most differentiate any target group.
Descriptors are the variables used for differentiating. Descriptors can include demographic, psychographic, or behavioral variables or some combination of all these. Learn more.
- Descriptor groups created by the Project Owner are available to all published users after the project is published.
- The descriptor manager displays descriptor groups under the "Owned by project."
- In the project tree, descriptor groups are displayed above the divider line at the top of the tree.
- Creators and Explorers can create their own descriptor groups.
- The descriptor manager displays descriptor groups under "Owned by me."
- In the project tree, descriptor groups are displayed at the top of the tree under the divider line.
3. Project tree glossary
In the project tree, variable types are represented by a corresponding icon. Fonts are also used to highlight the attention to specific items, for example, new items in the tree.
Some of these icons/colors are only visible to the project owner. Learn more about User Roles.
a. Project tree icon color
The color of the icons in the project tree indicates the commonality of items across multiple sources. The one exception is on constructed items, where items display green regardless of how many sources the root axes belong to.
A green icon indicates that the variable and responses within are common in all sources.
In this example, the axis “Length Of Visit” is common to the first and second sources, as the axis is green. Both source names are orange on the left side panel.
Constructed axes and elements display green regardless of how many sources the root axes belong to.
A blue icon indicates that the variable is not common in at least one source.
In this example, the axis “Q2b Age Groupings” is not only blue but is also bold as the variable is new to the project. The second source name is orange in the left side panel, as the highlighted item exists in the second source.
Green and Blue Icons
A green and blue icon indicates that the variable is common across all sources, but responses within the variable are not common across all sources.
In this example, the axis "How Likely Will You Be to Return To Hotel" is common to the first source and second source, but the elements "Don't Know" and "Missing Info" are not. The second source is orange, as the highlighted item exists in the second source. These elements are also bold as they are new to the project.
b. Project tree font and color
In the project tree, new items display in bold. New items result from:
- Adding or updating a data source in the project
- Or a user-created construction
When a display label is changed, i.e., the item is relabelled, the new label displays in italic font in the project tree.
In this example, the display label for the element “Auckland” (at the end of the axis) was changed to “Auckland National.”
In the project tree, invalid constructions are shown in red. Using the ERR icon under the Project tree filters, you can also identify invalid constructions. Learn more.
When auto sync is off (available to Project Owners only), changes are identified by a blue font. Variables that will change once syncing is on will show in blue.
c. Project tree icon description
Within your data design, there will be different sections of the survey you wish to organize into headings. Headings do not have an icon indication in the tree but can be automatically opened by clicking on the text or the arrowhead.
A standard axis is a discrete variable with elements for each response. An axis contains elements that group data records.
For market research data:
When colored green, it indicates the question and answers are common to all sources.
|Axis (not common to all sources)||
When a question from your survey does not exist in all sources.
|Axis (common but elements within not common to all sources)||When a question from your survey exists in all sources, however, some of the answers/elements within are not common to all sources.|
A category within a variable. A response to a question; an item in an axis.
When colored green, it is common to all sources.
|Element (not common)||When an answer from your survey does not exist in all sources.|
Orange dot - Axis constructed by the user using construct - new axis.
Constructed axes display green regardless of how many sources the root axes belong to
|Element (constructed)||Orange dot - Element constructed by the user using construct - new element.|
Yellow dot - Axis created by combining multiple questions into one new item using combine items, or by automation when merging single response questions into multiple response questions on creating a new project.
Combined axes display green regardless of how many sources the root axes belong to.
Yellow dot - Element created by combining multiple elements into one new item using combine items.
|Grid-Header||A grid contains a group of axes that have a common element list. Grids are created automatically based on the automation algorithm when creating a new project. The project owner can also create grids.|
|Grid-Group axes||Axes within a grid. These can be renamed or removed from the grid individually. The icons of the axes in the grid follow the green, blue, or green& blue color pattern according to their commonality across data sources.|
|Grid-Elements||The common element set within your grid that contais the elements from each axis within the grid. Labels within can be edited and combined.|
Elements combined within a grid display the orange C icon.
|Measure||Measures are used to analyze continuous variables. Measures contain numerical values. When a measure is applied to a table, you will see the count of the respondents in a cell and then the value-based response for those respondents.|
Measure group is another construction type available in Harmoni. Measure groups combine a group of measures that are logically linked together.
|Construction from a measure. A linked axis has a one-to-one relationship with the measure from which it was created.|
Elements within a linked axis. A linked element is created for each unique value in the measure the axis is linked to.
Verbatims contain raw text. Open-ended "verbatim" response asked in a survey. Dragging this to the analysis will display the text of the response and a graphical word cloud. Axes and measures can be added to the verbatim item in the analysis.
|Weight||Weight is a continuous variable with numerical values. Weighting measure that can be applied to the analysis.|
|Projection||Using Population Projection will allow you to view analysis results as population figures and see the actual size of the market.|
|Date||An axis can be flagged as a date axis using the data design options. This enables an offset reference to be applied and will automatically generate a line chart when choosing to visualize.|
|Display Errors (Red Font)||
Invalid constructions are shown in red in the project tree. You can also identify these by using the ERR icon under the Projet tree filters.
|Sort within (ascend)||The icon will appear when sorting items in ascending order within the selected item. Any new item (from an updated source/or added source) will appear in the applied order.|
|Sort within (descend)||The icon will appear when sorting items in descending order within the selected item. Any new item (from an updated source/or added source) will appear in the applied order.|
|Fix to top||When applying "fix to top," this icon will appear in the dictionary tree to indicate the item's fixed position.|
|Fix to bottom||When applying "fix to bottom," this icon will appear in the dictionary tree to indicate the item's fixed position.|
4. Reset Project tree
The reset tree option under the search or the design menu allows you to collapse the project tree (collapse all) easily.
Where to from here?