Symphony is an application that assists you with organizing, analysing, and reporting on themes found in text data. Generally speaking, any text that can be summarized into one or more thematic ideas is suited for use with Symphony. Examples include:
If you have had to organize texts by theme, you know how challenging it can be when you have hundreds or thousands of paragraphs. Some of the challenges:
Symphony is optimized for dealing with these issues. The basic idea is to free you from the administrative task of organizing data so you can focus more energy on the actual analysis - what is the data telling you.
The Symphony Content Analysis Console is the launch point and home page for the web-based Symphony. From there you open projects and invoke all Symphony features. The most frequently-used features are made accessible from a row of buttons at the top of the console. The remaining features are accessed through the main menu which is accessed by clicking the Menu button () in the upper-left corner.
Your browser must support JavaScript, Html-5, and CCS-3. Basically that's any of the popular modern browsers.
Google Chrome | Windows, Linux, iOS. | |
Microsoft IE (Edge) | Windows; uses Google Chromium engine, making it comparable to Chrome. | |
Firefox | Windows, Linux, iOS. Some things don't display correctly. | |
Safari | Seems less standards compliant than others; recommend Chrome instead. |
To use this application, you need a license, which is bound to your email address. If you don't have a license, you can purchase one by contacting sales@activejava.com. After a license has been issued to you, do the following.
Note that if your organization has a Single Sign On (SSO) arrangement with Active Java, the steps will be as defined by your organization.
Each project you create is an analysis initiative. Examples are 360 reviews, team assessments, employee engagement surveys, and Tweets containing a particular hashtag. Ultimately your objective is to uncover and measure themes in a body of text.
Symphony projects are stored in stand-alone files with an .ca extension.
To start a new project:
Once you've created or uploaded a Symphony project, you can manage its content and produce reports from it. As a minimum, this means engaging in the following tasks:
Depending on the nature of your project, each of these tasks might be performed multiple times and might utilize different Symphony features. The icons on the main toolbar toward the top of your browser give you access to the features that enable you to do most of your work. Additional features are available by clicking the Menu () button. Anytime while using Symphony you click the help () button or press function key F1, this help document will be loaded into your browser, preset to the topic that covers the view you have open.
To open a project:
The summary project information will appear to the right of the list. At this point, opening any views or forms will display data relevant to the project.
Once you've completed your analysis, or if you want to start an existing project from scratch, you can delete it from the server. It is advisable that you make a backup of the project (using the Download option) if you think you might change your mind at a later date.
If you decide to change the name of a project, do the following:
If you have a Symphony project file that is not already on the server, you can upload it. You might do this if you created a project with Symphony Desktop, or if you downloaded and deleted a project from the server and now want to add it back.
If you want to archive a project, or if you prefer to store your project on your computer instead of on the server, you can choose to download a copy to your computer.
Symphony in conjunction with the Symphony Synchronization Service can provide you and members of your organization the option to work collaboratively in real-time on the same project. For this feature to work, both you and your members must:
Anyone that fits these criteria will show up as being available for addition to your project
The Symphony Synchronization Service coordinates changes you and your team members make to ensure that everyone sees the project the same. When you make a change in Symphony, that change is also distributed to each of the team members that are connected to the project. If any members are using Symphony Desktop, they see the changes when they synchronize.
To give a team member access to your project:
When added, an automated email will be sent out to each new team member, instructing them on how to access the project. The instructions cover both the web and desktop versions of Symphony.
The Excel Import utility lets you import Excel spreadsheets into Symphony. Typical uses:
To import, you map the first respondent's data to Symphony. This is done by dragging cells from the spreadsheet to panels on the Excel Import form that are designated for various types of content. Symphony uses your cell mapping as a pattern to apply to each subsequent row. The pattern is repeated until Symphony encounters a row where the respondent name cell(s) have no value. At this point it is assumed that the end of the worksheet has been reached and processing stops.
When it comes to importing Excel spreadsheets, Symphony allows considerable flexibility with regard to format. Here's some guidelines:
To open the Excel Import form:
To upload an Excel spreadsheet to Symphony:
Symphony reads the spreadsheet and returns a sample of rows -- the first 40 rows -- from the first worksheet. Symphony also returns a list of worksheets, so if your data is not on the first one, you can switch. You will use the sample rows as a reference for telling Symphony how the worksheet maps to Symphony.
When importing an Excel worksheet to Symphony, you tell Symphony which columns you want to import, and what the columns contain. You do this by dragging cells from the worksheet into regions on the Import Excel form that have special meanings.
The key to mapping the worksheet to Symphony is to think this way:
What cells do you want to import for the FIRST RESPONDENT? This can be any combination of cells with values as well as header cells such as categories and questions. Symphony will apply that pattern to the remaining respondents.
Regardless of the nature of your data, you must include a cell or cells to uniquely indentify each respondent. This can be a number such as a response ID, or the respondent name if it can be guaranteed to be unique. If the worksheet doesn't include such a column or if the names aren't unique, you will need to add one. (A column with an auto-incremented value will do the trick.)
To import the respondent identifier:
If your worksheet has columns for respondent demographics, or if you have ratings items, you can have them imported. To Symphony, demographics and ratings are the same thing, in that both become ways to aggregate or filter your data by respondent characteristics.
When demographics or ratings are imported, Symphony saves them as "Tag Groups" and "Tags". For Example, a rating item from a survey is a Tag Group, and each rating value is a Tag inside the group.
To map columns to demographics:
If you are importing a survey, the comments are probably the main reason you are here. Symphony lets you import as many comment columns as your worksheet contains. You can also optionally import decorators -- codes, tags, and tag groups -- that are related to the comments.
To map a comment cell to Symphony:
You can have cells imported as codes, and have the comments initially coded to them. You might for example want to import the text of a question along with the comment. If you do this, a code will be created from the question and the comments that are in response to that question will be coded to it. This is a common practice; it provides an initial context for thematic coding. After all the comments have been coded to themes, the question code is empty and can be deleted.
Another practice is to import text that represents a high-level category for comments that are grouped together in the survey. For example, let's say the survey asks three questions relating to "Strategy". If your worksheet contains a cell with "Strategy" in it, you can map the three comments to it and they will all be coded to "Strategy" during the import.
To map excel code columns to Symphony:
You can optionally import Tag Groups and Tags along with comments. This is in addition to codes. This feature results in the imported tags being attached to the corresponding comments. This is not a common practice, but it is available for associating additional permanent or temporary attributes with individual comments. You could for example attach the text of a survey question to each comment as a means of keeping a permanent record of the question the comment was in response to.
To import tags and tag groups, do the following:
When you are satisfied with your Excel - to - Symphony mapping of content, click the Run () button. Your map will be uploaded and applied to the worksheet. A message will be displayed when it is finished. You can now close the form. If you go to People view or Coding view, you will see the comments and codes that were imported.
The Word Import utility is used to import interview transcripts that have been stored as Word documents. As a minimum, the document will contain the verbatim comments you will attribute to a single interviewee (respondent). It can however contain other text. For example, it is a commom practice for those who use an interview guide to transcribe the comments into a copy of the guide and them import the copy. Symphony can be directed to import only the texts that belong to the interviewee, or alternatively additional texts such as the actual questions can be imported and used as an initial code and/or a tag for the comments.
The extraction of text is based on Word Styles. So it is important that your document be consist with styles usage. If you are not familiar with Word styles, you will find an abundance of explanations if you Google it.
To import an interview transcript saved as a Word file:
The styles-to-content mapping is how you tell Symphony which content to import and what to do with it. The styles are the Word styles used by the texts in your document, the content-types are the various types of Entities you can create from the text.
When a style is mapped to a content type, an icon appears to the left of each paragraph that uses that style.
To map word styles to content:
The Text Paragraphs list shows all the paragraphs of text in your document. While doing the styles-to-content mapping, you might find it useful to hide text that is irrelevant and display only the text you are interested in. For example, if your document has a lot of boiler-plate text. You can do this by choosing to ignore all the text that uses a given style.
While previewing the document, you may see that you have content that is incorrectly styled. For example, you might see a comment that is not using the style mapped to comments. If you leave it as-is, it will not be imported as a comment. You can fix it as follows:
When you are finished mapping Word styles to Symphony content, you are ready to perform the import. To do this, click the Run () button. A message will display when the import is finished. You can now close the form. If you go to People view or Coding view, you will see the comments and codes that were imported.
This feature enables you to import a text file. The entire contents of the file is imported as Comments, each one attributed to the Respondent you choose. A separate Comment is created for each paragraph. Example Uses:
To import a text file:
The Transcripts feature is used for converting 1) audio recordings from interviews, and 2) Word documents from transcription services into Symphony content. The following capabilities are provided in support of these objectives:
The following topics provide guidance on how to utilize these functions.
This option provides you the capability to upload an audio file to your project. You would do this if you want to include text from the audio file as codes or comments in your project. After the file is uploaded, the transcription process begins immediately. You'll receive an email when the transcript is ready. It takes about 20 minutes to transcribe one hour of audio.
To upload an audio file for transcription:
This option is for uploading transcripts in Word files, typically what you receive from a transcription service. It is also used to upload a transcript that was previously downloaded here and edited in Word. The file must conform to this format:
Symphony attempts to match speaker names up with names of team members and interviewees. If matches don't exist, you can set them manually.
To upload a transcript:
After the upload is finished, the transcript will open for editing automatically.
The primary purpose of editing the transcript is to prepare content to be imported as codes or comments.
A button is provided on the toolbar above the transcript for each speaker. If the transcript was produced by uploading an audio file, the speakers will not be associated with team members or interviewees so you will need to set them. If the transcript was uploaded from a Word file, Symphony will make an attempt to match the names in the transcript with names in the projects. If matches are not found, you will need to set them.
To make it easy to identify speakers with text in the transcript, the text and the speaker buttons are color-coded.
If a paragraph is attributed to the wrong speaker, right-click the paragraph then choose the correct speaker from the context menu.
You can add new codes to your coding structure by converting team member speaker paragraphs codes. How you approach this depends on whether you want to create a new code in your coding structure, or if the code you want already exists. Once a paragraph is turned into a code, a folder icon appears next to it. When you click the import button, all comments that follow it will be coded to it until another code is encountered.
Approach 1
Approach 2
All the texts of speakers designated as interviewees is considered to be comments, each paragraph being a separate comment. However they are not imported into your content as comments unless you explicitly import them. This can be done manually one-by-one, or en mass using the Run Import option described below.
To manually code a comment:
To run the automatic import:
After you've finished with a transcript, delete it as follows:
People View is for managing respondents and their comments. In general, it is more efficient to import respondents from a spreadsheet and their comments either from a document or a survey spreadsheet, but this view provides the capability to add then one-by-on as well as edit their information after they've been added to the project through other means.
People View is organized into four sections:
In most cases, Symphony refers to people as "respondents". Comments are attributed to respondents. A respondent does not have to be a person; it can also be an event such as a focus group. Regardless of the actual entity, comments are attributed to respondents, and any attributes such as demographics that belong to the respondent are inherited by the comments attributed to the respondent.
To open People View, click the People button on the console toolbar.
It is usually quicker to import your respondents from a spreadsheet. You can however add them manually as follows:
Demographics are custom attributes pertaining to respondents. Examples are Gender, Location, and Tenure. Each demographic has "demographic values" such as Male and Female for Gender.
The primary purpose of demographics in Symphony is to provide ways of grouping and/or filtering respondents to perform deep analysis or targeted reports. For example, with a Gender demographic you can compare the average comment sentiment for Males and Females. Or run a Feedback Report that contains comments belonging to respondents from a particular country.
Symphony implements demographics as "tag groups" and "tags". A demographic such as Gender would be a tag group, and Male and Female would be tags within the group. What differentiates a demographic from other tag groups is that you define the tag group to be used for respondents only.
To manage the demographic values assigned to a respondent:
Generally if your comments exist in electronic form, it is more efficient to import them using the Excel or Word Import utilities. However, you can add/edit comments for a respondent directly in People View.
You can add comments for a respondent using your computer's clipboard. Any text that is copied to your computer's clipboard can be pasted as new comments for a respondent. Each paragraph is saved as a separate comment, and they are placed inside the Unassigned code.'
Code View has additional features to help you code efficiently. However, if you want to focus on coding comments belonging to specific respondents, you will find People view more suitable since the comments are organized by respondent.
The ideal comment consists of a single theme relevant to the project. This is so that only the data relevant to a report or analysis is included in filter criteria. For projects such as large employee engagement surveys, this might not be practical, in which case you may want to take advantage of Symphony's multi-codeAllows comments to be coded to more than one code feature. However, for projects such as team assessments or comparisons of job candidates, single-themed comments might add to the overall project quality. Symphony's split comment feature gives you a way to do this efficiently. When you split a comment, Symphony removes the split portion from one comment and creates a new comment with the split portion, and attributes it to the same respondent. The split portion inherits all tags attached to the original comment, but these can be managed separately without affecting the original.
Comments can be split in People view and Coding view. They can also be split directly inside the Comment Properties form any place this form is available.
To split a comment in the Content Editor:
To split a comment in the Comment Properties Form:
Merge Comments is the opposite of splitting: you want to combine two comments into one. The result is the same as if you select the text of one comment, delete the comment, then paste it to the end of another comment. In addition, if the consumed comment has tags that the consuming comment doesn't have, they are added as well.
The text from the dragged comment is appended to the text of the target comment as a new paragraph, and then the comment is deleted. Any tags the dragged comment had are added to the target comment.
Field notes are basically comments about codes or comments. They are intended to provide you and your team with a way of communicating something about a comment, usually where you want to revisit it later on. For example, you might want to leave a field note asking a team member to reconsider the way they have a comment coded.
Field notes are visible in the comment's properties form. In Review view, an additional icon is displayed next to comments to indicate that they have field notes.
To create or view a field note for a comment:
Symphony provides an efficient way of editing respondents, comments, and codes while navigating around People view.
Symphony's Coding view provides the capability to manage your coding structure and the location of comments within it.
Coding View is organized into three main areas:
Coding Structure - Displays your coding structure as a hierarchy tree | Content List - Displays the codes and comments contained inside the code that is highlighted in the coding structure |
Content Editor - Displays the text of the code or comment selected in the Content List. You can directly modify the text here |
To open Coding View, click the Coding button on the console toolbar.
New codes are always created either as a top-level code within the coding structure, or as a sub-code of an existing code. These instructions describe using the coding structure's context menu, which is the easiest way to create a code while engaged in coding activities.
This can be done one of two ways:
To move a code or comment up or down amongst its peers:
A code or comment will move only if it can. For example, if a comment is the last in a code, clicking the down arrow will have no effect.
Depending on your objective, there different actions. These actions can be modified with the Alt Key.
Codes can be deleted only when they contain no codes or comments.
Sometimes you might want to move the comments out of one code and into another and delete the code. You can do this by selecting the comments, dragging them into the destination code, then delete the code.
Symphony provides a short-cut approach that moves the comments and deletes the code all in one motion. The action also applies to sub-codes and their contained comments. So essentially you can consolidate an entire tree of codes and comments into a single code.
To edit the text of a comment:
Depending on your analysis objectives, you may want to split a comment into two or more separate comments when it contains more than one theme. When a comment is split, each portion is tracked as a separate comment attributed to the original respondent, and has its own coding and tags.
To split a comment:
Symphony provides the capability to make field notes for individual codes and comments. An example of a field note for a code is an elaboration on what you had in mind when you created the code. An example of a field note for a comment is you suggest to your team mates that the comment should be moved to a different code. You can use the code and comment field notes for whatever purpose you desire. But the general intent is to provide a way of making a statement or bringing attention to a code or comment.
Field notes are recorded in the lower section of the code and comment properties forms. To make a field note or to read one, simply double-click the code or comment to bring up its properties form.
You can optionally display a statistic next to each code in the coding structure. This can be useful during analysis to gain a sense of comment distribution across your codes. Values are updated automatically as changes are make to your project. Here are the choices:
To set the display statistic:
Symphony provides an efficient way of codes and comments while navigating around the Coding tab.
The Word Tree organizes the words in your project into a hierarchy of "nodes", where each node represents one or more words found near each other in a set of comments. The frequency the word/phrase appears in your data is included. The primary use of the Word Tree is to uncover themes. Unlike Symphony's Cluster Analysis, the Word Tree can identify multiple themes in each comment. This makes it ideal for projects where multi-themed comments are common (e.g. employee engagement surveys). For other types of projects, it can still be used but you should keep the "Auto-Run on Code" option checked so as to unintentionally moving comments from one code to another repeatedly.
The Word Tree page is organized into four sections:
To open the Word Tree View, click the Word Tree button on the console toolbar.
A big part of using the Word Tree is taking the time to understand what from the result is useful to you. Your data is decomposed multiple times into different branches. This can be a bit overwhelming, but gives you the latitude to pick a starting point that seems important to you analysis. Here are the things that will help you zero in on the nodes of interest:
The Word Tree page includes a copy of your coding structure so you can code comments as well as maintain your coding structure.
You can create a new code from a Word Tree node and have its Comments coded as follows:
A Venn Diagram is made up of circles representing data. The circles overlap according to how closely the data is related.
In Symphony's Word Tree view, the circles represent words or phrases found in comments. The diameter of each circle is determined by how often the word or phrase appears.
To create a venn diagram, check the checkbox next to each word or phrase you want to include.
To remove a word or phrase from a venn diagram, click its checkbox to remove the checkmark;
You can save a venn diagram to your computer by clicking the Save Diagram button. The diagram is saved as a .svg file. You can open it directly in a browser, or you can insert it into other applications that accept .svg files as images.
Generally speaking, when you perform actions such as dragging a Word Tree branch onto your coding structure (which creates a code and moves the branch's comments into the new code) it is a good idea to re-run the analysis with the new code filtered out. The reason for this is that performing additional coding may have unintended results, particularly if your project's Multi-Code setting is turned off.
To make this easy, Symphony includes a checkbox setting that does this for you after new codes are created. By default it is checked, meaning after a code is created a filter will be set and the analysis will run again. If you don't want this behavior, uncheck the box.
The Symphony Review is intended to enable you to get a feel of the overall flow of your analyzed content toward the end of your coding efforts. At the same time, it provides the same capabilities for creating and editing codes and comments and changing their positions as does Coding view. The difference is that Report view is optimised for viewing as opposed for coding. It provides a continuous flow through your coding structure, the same as it would be for a printed report.
To open Symphony's Review, click the Review button on the console toolbar.
In Symphony's Review view, the easiest way to code a comment is as follows:
The order of codes and comments matters in that how you see them here in Review view is the how you see them in a Feedback report.
To move a code or comment up or down amongst its peers:
A code or comment will move only if it can. For example, if a comment is the last in a code, clicking the down arrow will have no effect.
To modify the text of a code or comment:
The code and comment properties forms include an area for entering a field note about the code or comment. This can be used for any purpose you desire, such as a reminder to yourself about next steps related to the code or comment, or instructions to your colleagues. To indicate that a code or comment has field note, the Review displays an icon next to it.
To add or view a field note, double-click the code or comment to bring up its properties form. Also, you can quickly view field note by moving your mouse over the icon.
Decorators are additional pieces of information optionally displayed about a code or comment. For codes, decorators are a statistic. For comments, decorators are either tags such as demographics or the name or title of the respondent that the comment is attributed to.
To choose decorators:
Symphony's Queries view enables you to define and run queries against your project. Queries consist of:
Symphony's Queries view is quite powerful and quick. It makes it possible for you to ask pretty much anything of your data and get an almost immediate response. If your project includes demographics or ratings items defined as demographics, you should find that Symphony's Queries view provides you with the tools you need to gain deep insights into your data.
To open Symphony's Query Utility, click the Queries button on the console toolbar.
To define a new query:
Option | Description |
Hierarchy | The query rows are arranged into a drill-down hierarchy. On the Feedback Report form you map each query column to a Word style in your template; this determines how the text will be formatted. |
Table | The query results are arranged into a Word table defined in your Word template. |
Once you've chosen columns for your query, you can customize various properties for each. Do this by double-clicking a column header or right-clicking the header and choosing Properties to bring up the Query Column Properties form.
Property | Description |
Caption | This is the displayed in the header row of the output. By default this is the name of the column. |
Value Required | This has different meanings for different types of columns. |
Min Value | If you want to restrict output to rows that meet a minimum value, enter it here. |
Max Value | If you want to restrict output to rows that don't exceed a certain maximum value, enter it here. |
Hidden | When checked, the column is not included in the output. This option is used when you want to use a column to place restrictions on rows but don't want the column included in the results. |
Suppress Repeats | This applies to Entity and Attribute columns only. When checked, the value for the cell is displayed only when the underlying Entity changes. Otherwise an empty cell is produced. |
Harvey ball | This column applies to Statistic columns only. When checked, this column displays a Harvey ball instead of the statistic. The Harvey ball to display is determined by the "upper limit" setting for each Harvey ball. (These settings are visible only when the Harvey ball box is checked.)
When Harvey balls are used in a crosstab query: the Harvey ball upper limit settings apply to the individual crosstab columns. Tip Not sure what values you want for the Harvey ball upper limits? Run the query once with the Harvey ball setting unchecked, take note of the values in the query results, and use what you learned as the basis for the values. |
Pre-Sort | The pre-sort sorts the query results immediately after the query executes and BEFORE the Max Rows setting is evaluated. This gives you an opportunity to have the desired rows returned in the event that Max Rows is set. For example, if you want the rows where a particular statistic value is the highest, you can have the results pre-sorted in descending order on that statistic; the rows with the highest values will survive the Max Rows test, if it is set. |
Post-Sort | This determines the final order of the rows returned. Choose the direction from the dropdown. The results can be sorted on multiple columns, in which case the sorts are performed from left to right first on Entities, then Statistics. |
A crosstab query is a query that calculates a statistic, and then groups the results by two sets of values — one for each row and the other for each column.
Here's a query result that is NOT a crosstab:
Organization | Region | # of Respondents |
Customer Service | North America | 1025 |
Customer Service | Asia | 546 |
Customer Service | Europe | 438 |
Marketing | North America | 687 |
Marketing | Asia | 138 |
Marketing | Europe | 97 |
Technology | North America | 254 |
Technology | Asia | 94 |
Technology | Europe | 408 |
Operations | North America | 2659 |
Operations | Asia | 869 |
Operations | Europe | 482 |
And here's the same query with the Region column displayed in a crosstab:
Organization | North America | Asia | Europe | Total Respondents |
Customer Service | 1025 | 546 | 438 | 2009 |
Marketing | 687 | 138 | 97 | 922 |
Technology | 254 | 94 | 60 | 408 |
Operations | 2659 | 869 | 482 | 4010 |
Crosstab queries are created by converting an Entity column in a existing query into a crosstab column. Do this as follows:
Here are some important things to know about crosstabs:
After you run a query, you can download it to an Excel file as follows:
If you decide to delete a query you defined, highlight it in the Queries list and click the Delete () button directly above the list.
In addition to the queries you define, Symphony also creates four queries in every project. These queries can be modified, but they cannot be deleted. Here's what they each do by default:
Queries View includes four lists of columns that can be added to queries. This section describes each.
Entity Columns represent the content in your project such as codes, tags, respondents, and so forth. Entities normally have relationships with other entities -- often many relationships. For example, a code has a relationship with all the codes and comments inside of it. When an entity column is included in a query, it effectively function as a row-level filter. What this means is that a row is output for each entity. If multiple entity columns are included in a query, a row is output for each combination of entities.
Column | Description |
Project Member Column | Outputs the name of each Symphony user with access to your project. As a minimum it includes you but can include other users if you have a shared account and you give them access. |
Respondents | Outputs the name of each survey respondent or interviewee for your project Respondents are the entities to whom comments are attributed to. |
Code Level 1, Code Level 2, etc | Outputs your codes at the respective levels. |
Comment | This column produces a row for each comment, consisting of its text. |
Demographics | If you have any demographics defined, they are listed next. Including one of them outputs the values of each of the demographic values. For example, a Gender demographic might output two rows: one for Male and one for Female. |
Tag Groups | Outputs a row for each Tag Group in your project. Tag Groups include demographics because demographics are set up as a special kind of Tag Group. |
Tags | Outputs the Tags contained inside your Tag Groups. |
Word/Phrase | Output each word/phrase found in the project. |
Statistic Columns produce numbers for you. They can be used by themselves, but when they are combined with one or more entity columns, a row is produced for each entity combination and the statistic is based on the entity.
Column | Description |
# of Respondents | This gives you the number of unique respondents in each row. |
# of Comments | This gives you the number of comments represented in each row. |
# of Words | This gives you the total count of words contained in the comments represented in each row. |
Phrase Count | This is used in conjunction with the Words/Phrases Entity column. It gives you a count of how many times each unique Word/Phrase appears in each row. |
Avg # of Words | This gives you the average number of words appearing in the comments represented in each row. |
# of Respondents Comments | This gives you the total number of comments belonging to the respondents represented in each row. |
Demographic Size | This gives you the number of respondents belonging to the demographic(s) included in the query. If no demographic columns are included, this is always equal to the number of respondents in the project. |
# of Respondents - Overall | This is the number of unique respondents included in the query. |
% of Demographic Value | For each row, this is the percent of unique respondents from the demographic(s) included in the query. |
% of Crosstab | For crosstab queries. |
% of Respondents | For each row, this gives you the percent of respondents included in the query. If the query doesn't included a filter, the percentage is based on the total unique respondents in the project. |
% of Respondents - Overall | This gives you the percent of respondents in the project represented in each row. |
Demographic % of Respondents | For each demographic value, this gives you the percent of respondents in the query. |
% of Comments | For each row, this gives you the percent of comments included in the query. If there is no filter, this is based on the number of comments in the project. |
% of Comments - Overall | For each row, this gives you the percent of comments included in the project. |
% of Demographic Comments | For each demographic represented in the current row, this gives you the percent of comments belonging to the underlying respondents. |
Positive Sentiment | This gives you the average positive sentiment associated with the underlying comments. The average includes comments whether they have positive sentiment or not. |
Positive Sentiment - Strict | This is the same as the Positive Sentiment, but the calculation includes only comments that have positive sentiment. |
Negative Sentiment | This gives you the average negative sentiment associated with the underlying comments. The average includes comments whether they have negative sentiment or not. |
Negative Sentiment- Strict | This is the same as the Positive Sentiment, but the calculation includes only comments that have negative sentiment. |
Net Sentiment | This is the average sentiment. If the value is more than .5, the overall sentiment is positive. |
Count of Level 1 Codes | This gives you the number of top level codes used by the comments represented in each row. |
Count of Level 2 Codes | This gives you the number of second level codes used by the comments represented in each row. |
Count of Level 3 Codes | This gives you the number of third level codes used by the comments represented in each row. |
Count of Level 4 Codes | This gives you the number of forth level codes used by the comments represented in each row. |
# of Tags | This is the number of tags used by the codes, comments, and/or respondents represented in each row. |
% of Tags | This is the number of tags used by the codes, comments, and/or respondents represented in each row presented as a percentage of the total tags. |
Count of Tagged Content | This is the count of entities in the current row that have tags associated with them. |
Generally speaking, these columns provide additional information about an Entity (code, comment, respondent, etc.). These are similar to the Entity columns, but they give you control over what is being displayed for the entity.
Column | Description |
Comment ID | This column produces a row for each comment, consisting of Symphony's internal comment ID assigned to the comment. |
Comment Field Note | This column produces a row for each comment, consisting of the comments's field notes. |
Code Text | This column produces a row for each code irrespective of code level, consisting of the text of the code. |
Discussion 1 | This column produces a row for each top level code, consisting of the notes for the code. |
Discussion 2 | This column produces a row for each second level code, consisting of the notes for the code. |
Discussion 3 | This column produces a row for each third level code, consisting of the notes for the code. |
Respondent Title | This column produces a row for each respondent, consisting of the respondent's title. |
Respondent First Name | This column produces a row for each respondent, consisting of the respondent's first name. |
Respondent Last Name | This column produces a row for each respondent, consisting of the respondent's last name. |
Outline | This column produces a row for each comment, consisting of a dot-separated number that shows the coding hierarchy path to the comment. |
Charts you've defined in Charts View can be included as query columns. This results in a chart being generated in the corresponding cell for each row. The data for the chart comes from the query assigned to the chart, and the data for the chart's query is the subset of data for the row the chart is being generated in.
With chart columns, you can:
You can set a filter to narrow the scope of content included in a query. For example, you can have a query return comments only for respondents belonging to a particular demographic.
To set a filter:
Symphony's Spawn a Chart feature is a short-cut for creating a chart. After you define a query, you can simply tell Symphony to spawn a chart, and the new chart that uses the query will be available immediately in Charts view.
To spawn a chart:
With Symphony Charts you can produce charts of your data. The charts use data from queries you define in Queries view. The charts are displayed in real-time from the queries, and can be included in Feedback reports.
With Symphony Charts, you can
As a minimum, all charts require the query to include at least one Statistic column. A single Entity or Attribute column can be included to produce a drill-down. If extra columns are included, they are ignored. Exceptions to the data requirements are noted in the table below.
Chart | Description | Query data requires |
Column | Vertical bars where the height of each is determined by an assigned value | |
Bar | Horizontal bars where the eight of each is determined by an assigned value | |
Pie | Circle divided into slices, where each the size of each slice is an Entity whose size is determined by an assigned value | The pie is sliced using the first visible Statistic column in the query; subsequent Statistic columns are ignored. |
Line | Series of line segments connecting Entities represented by dots, the vertical position if each determined by its value | |
Scatter | Displays Entities in an XY grid |
|
Bubble | Same as Scatter, but a third statistic is used to define the size of each data point. |
|
Radar | Displays Entities in a circular grid, where the distance of each entity from the center is a measure of the relationship strength to another Entity. |
|
Treemap | Displays Entities in a rectangular grid, where the space occupied by each is a percentage of an aggregate statistic. Ideal for presenting the relative predominance of items in a hierarchy (e.g. how many Respondents are coded to each code throughout the coding structure) |
|
To open Symphony's Charts Utility, click the Charts button on the main toolbar.
To create a new chart:
The data used by charts is provided by queries defined in Symphony's Queries View. By default, new charts are set up to use the predefined "Sandbox" query. You can change this to any query in your project.
To choose the type of chart you want to create:
The query used by a chart is run automatically when you make a change. If however you modify the query itself -- for example you switch to Queries view and add or remove a column -- the query will not run. To force it to run, click the Run Query button.
To resize the chart, either
Symphony's Clone a Chart feature creates a new chart from an existing one. You can for example experiment with the built-in Sandbox chart, then when you have something you want to keep, clone it.
To Clone a Chart:
After you've generated a chart, you can have it saved to your computer as a .PNG file. The dimensions of the file are the same as the chart region.
Symphony Word Cloud Generator creates a word cloud from text your project. A word cloud is an image that has words arranged in a way that conveys their relative frequencies. There is a variety of ways to depict word clouds. Symphony does it by varying the font size. Words are arranged in alphabetical order for ease of finding a specific word.
To open Symphony's Word Cloud generator click the Word Cloud button on the console toolbar.
All settings for a word cloud are optional. You can simply click the Create Word Cloud button.
This setting lets you control how large the word cloud is in terms of pixels. If you don't have much text you might want to make it smaller. Or if you have a document you want to include the word cloud in, you might want to set is to a size ideal for the document. These instructions assume that you have just generated a word cloud and it is currently being displayed.
Using your mouse, drag the lower-right corner of the word cloud region until it is sized as desired. As you drag, the current width and height in pixels is displayed above the region.
After you generate a word cloud, you may find words that you don't want included. You can exclude them by adding them blacklisting them. Symphony automatically excludes some words such as "the" and "is". If you see additional words that you don't want included, you can add them to the blacklist. When you update the word cloud, it is generated without those words.
This setting lets you specify a word that you want to make the subject of the analysis. When you specify a filter word, Symphony uses only paragraphs from your text that have that word in it. Paragraphs that don't have the word are ignored.
This setting works in conjunction with the Filter Word field; it has no effect on the result unless a Filter Word is specified. the Word Proximity is an additional filtering setting. It limits the words included in the analysis to those within the Proximity setting of the Filter Word. So only words that are within the proximity of the Filter Word are included in the analysis. If the Proximity is zero, the entire text of each paragraph containing the Filter Word is used.
You can save your word cloud to your computer. The file is saved as a .PNG image file. Simply click the "Save Word Cloud to .PNG file" button.
Symphony's Feedback Report option produces a Microsoft Word report from your project data. By default, the report consists of your coding structure and comments organized into a hierarchy around the coding structure. It can also include tables of data from Symphony Queries and charts from Symphony Charts.
The report is produced from a Word document that you provide as a template. As such, the report has the look and feel that you provide. You tell Symphony what content you want included in the report by placing Word Bookmarks where you want output from Symphony to start.
To open Symphony's Feedback Report Utility, click the Feedback Report button on the console toolbar.
A feedback report template is a Word document to which Symphony outputs the feedback report content -- datasets produced from queries defined in Queries view and charts defined in Charts view. The template can consist of anything from a simple blank Word document to something much more complex that contains your organization's branding, boiler-plate text about your analysis methodology, and a summary of your analysis findings or other information related to your project but not in the scope of Symphony.
The data for a Feedback report is provided by queries defined in your project and accessed in Queries view. You can have data from any number of queries output to the report. For example, you can include a table of Respondents, tables of various slices of demographics and statistics, and your coding structure with supporting comments all in one report. You can also include charts from Charts view driven by the queries.
To assist you with getting started, Symphony includes a template called "Symphony Default Feedback Template.docx". You can use it to get a sense of how Symphony Feedback reports work. You can also download a copy of it and make modifications, then upload the revision to see their effects.
To include the results of a query in a report, you create a Word bookmark in your template where you want the output to start. (See your particular version of Microsoft Word on how to insert a bookmark.) Symphony looks for bookmarks with the same names as queries in your project, with these qualifications:
Each query has a "layout" setting, which instructs Symphony on how to process the query results.
If your template contains no bookmarks that match queries or charts in your project, Symphony will run the predefined "Default" query and will output starting at the end of your template.
You can have query results output to a table in a Word document. To do this, you add a Word bookmark in your template where you want the results to go. If you place the bookmark in a table cell, the results will start there and a row will be added to the table for each row from the query. If the bookmark is not inside a table, Symphony will create a table and move the bookmark inside it. To add a table:
To include a chart in your report, you create a Word bookmark in your template where you want the results to go. (See your particular version of Microsoft Word on how to insert a bookmark.) Symphony looks for bookmarks with the same names as queries in your project, with these qualifications:
Symphony makes this easy for you:
A Table of Contents (TOC) is a Microsoft Word field that is dynamically built from the contents of the document. What goes in the TOC is determined by the Word styles used in the document. Whenever Word encounters a paragraph of text that uses a style specified in the TOC field, it adds a corresponding entry to the TOC.
Symphony lets you maintain a collection of templates. Templates can be used on each project, or you can have templates that are customized for specific projects. This can be any .DOCX file, but ideally should be one that is designed for your project, including boiler plate text and images, your organization's headers and footers, and so forth.
Symphony provides you with a starter template that you can use as-is or customize to look the way you choose.
To upload a feedback report template:
The Symphony Default Feedback Template plus any you upload appear in the Word Templates list. The first step in generating a report is to specify which template you want to use. Once a template is chosen, the styles it contains are loaded into the Word Styles list. Symphony also makes an attempt to map content columns to styles.
Queries that use the Hierarchy format rely on Word styles for their presentation. If you are not familiar with Word Styles, a quick Google search will yield you more than you ever want to know.
When an Output Field is mapped to a Word style, a checkmark appears next to it. If you click on the Output Field, the style it is mapped to is highlighted and its radio button set.
To map an output field to a Word Style:
You may not need to map all the columns. For example, if your report only includes two levels of coding, you don't need to map styles for field that never get used. If while generating the report Symphony encounters a field that needs a style, it will tell you so can fix it.
You can optionally set a filter to limit the scope of data output to the report.
Decorators are additional pieces of information attached to codes or comments and are displayed along with it. In the case of codes, a predefined statistic can be used as a decorator. In the case of comments, decorators can be the name or title of the owning Respondent, or Tags (e.g. demographics) associated with the Respondent or comment.
A code can be decorated by at most one statistic. Comments can be decorated by as many items as are chosen, each separated by a comma.
To add decorators:
The look and feel of Feedback Reports is provided by the Word styles you use in your template. One thing that can be tricky is getting the comment numbers right if the style you use for comments includes them.
To make this easier, Symphony makes available an override that will restart the numbering in each code. When used, this option supercedes the bullet/numbering settings for your comment style.
To force comment numbers to restart:
Word styles can be defined such that a new page is started each time a style is encountered. This is useful for large sections of documents. Many Symphony users like to have a new page started for each top level code.
Quite often, Symphony users want the first occurance of a top level code to be treated differently. An example would be when your report template includes text such as a heading or explanatory paragraph that you want to appear right before the first code is output to the report. If the style used for top level codes specifies to start a new page, then your codes and comments would start on the next page, leaving your heading and/or explanatory text orphaned on a page by itself.
Symphony provides a workaround, which is to start new pages for you as the report is generated -- except for the first code. In turn, you omit the start new page setting from your Word style and the result is as expected: the first top level code is output where it is should be, and thereafter a new page is started for each of the remaining top level codes.
To get this effect:
The Report Distribution panel has settings for determining what to do with the report. To choose where to send the report:
To run the report, click the Run () button and wait for your report.
When the report finishes, it will be downloaded to your computer. Follow your browser's instructions for saving and opening the file. If you have fields in your report (e.g. a Table of Contents) Word will prompt you to see if you want to update it. The message itself is somewhat misleading in that it suggests that there may be external documents that it is dependent on. Ignore the meaning of the prompt and choose "Yes".
You can use a Feedback Report as a template for generating subsequent reports. In any given situation, this capability is useful for one of two reasons:
You don't have to worry about generated content in the report; when you upload the file, Symphony will remove it.
To use a report as a template:
Symphony's Filter Group lets you limit certain output on a variety of criteria. It is available in the following views/forms:
When a filter is set, subsequent actions in that view/form behave as if your project consists only of data that fits the filter. Filters can be set on codes, tags, respondents, words, and sentiment.
When filters are set on items within one of these broad categories, the results are broader with each additional item. For example, setting a filter on two codes results in output related to either of the two codes.
When filters are set on more than one of these broad categories, the results are narrowed. For example if a filter is set on codes and tags, the results consist of the content related to the subset of filtered code further limited by the content that is related to the filtered tags.
Where the Filter Group is available, a button is provided that looks like this:
. Clicking the button toggles the Filter Group open and closed.When a filter is set on one or more codes, only content that is related to the included codes is included in the results. For example:
All codes that are checked are included in the filter.
To set a filter on one or more codes:
When a filter is set of one or more tags, only content related to the included tags is included in the results. For example:
All tags or tag groups that are checked are included in the filter. Checking/unchecking a tag group causes all tags in that tag group as well as tag groups contain in that tag group to become checked/unchecked.
To set a filter on tags:
When a filter is set on one or more respondents, only content related to the included respondents is included in the results. For example:
To set a respondent filter:
You can limit filter results to comments and their related entities by the words contained in the comments, or by the size of comments. For example:
To set a word filter:
The sentiment filter lets you filter output based on comment sentiment. For example:
The sentiment filter consists of a group of radio buttons whose collective settings capture your intent. For Positive, Negative, and Neutral sentiment, you specify what to include as follows:
All settings are assessed for their impact but where they become mutually exclusive, the Neutral settings take precedence.
To set a sentiment filter:
The Miscellaneous tab currently has one setting: Max Comments per Code.
The Max Comments per Code settings isn't a filter attribute in the sense of the other Filter Group settings, all of which when applied to the project produces a copy of the project that consists of only content that passes the filter. This attribute instead functions as a marker that tells a query to stop outputting rows that rely on comments. It applies only in scenarios where comments are output. A typical application would be when you want to produce a Feedback Report that prints the text of comments, but you only want a small number of "representative" comments in the report. Any statistics that rely on the comments are unaffected.
By default, the Max Comments per Code is set to All, meaning the setting has no effect. To set a limit, you choose a value from the dropdown. A value of zero results in all other output being produced but no comments included.
If you have content in one language but need it in another, Symphony can translate it for you. Symphony accesses Google's Translation API on your behalf, translating text from virtually any language to any language. Google detects the source language, and translates it to the language of your choice.
The Language Translation tool is opened by clicking the icon at the bottom of the Symphony screen. Clicking it more than once toggles it in and out of view.
The tool displays the quota balance on your account, the number of characters you will be charged to translate the current scope, and the projected quota balance.
Translation scope is based on what content is selected:
Selection | Scope |
A comment | The comment you clicked on |
A group of comments (e.g. a selection range inside a code or belonging to a respondent) | The selected comments |
A Respondent in People view | All the comments belonging to the Respondent |
A group of respondents (by selecting a range) | All the comments belonging to the selected respondents |
A Code | The comments inside the code and sub-codes. |
When the scope is changed the calculated usage is updated in the Language Translation tool.
To start the translation:
Every Symphony license and license renewal comes with a quota of 50,000 characters -- roughly 15 single-spaced typed pages. If you need more, you can add to your quota by purchasing additional characters in blocks of 10,000.
Tags are custom attributes you create for content. They can be associated with codes, comments, respondents, or any combination of these, and can serve a wide variety of purposes.
Tags always exist in Tag Groups. Tag Groups are merely containers that identify one or more tags as having something in common. Gender as a tag group for example would have Male and Female as tags. A tag group called Sentiment might have several tags, each one representing a different measure.
In addition to containing Tags, Tag Groups can contain other Tag Groups, so it is possible to define any level of complexity you need.
Tags are available in all views via a "Tags Widget", which floats above the view. When you click on a content item in a view, it becomes the "focus" of the Tag Widget, resulting in the attached tags to be highlighted. Tags can be attached to and detached from the focused item by clicking the radio button or checkbox next to the tag.
To open the Tags Widget, click the Tags () button at the bottom of the page. Hide it by clicking a second time.
After Tag Groups have been created, you can change their attributes through the Tag Group Properties form.
If you no longer need a particular tag group or if you want to get a fresh start with it, you can delete it. Deleting a tag group also deletes all the tags and tag groups inside of it, as well as all tag attachments to entities, so be sure this is what you want.
To delete a tag group:
By default, tag groups appear in the Tags Widget in the order they are created. The order of the tag groups matters in two cases:
To create a tag:
To tag content:
To delete a tag:
You can move a tag up or down amongst its peers, or you can move a tag to another tag group.
To move a tag up or down:
You may want to move a tag from one group to another for a variety of reasons. For example, it is possible to define a tag in one group when you intended to define it in another. When you move a tag to another group, all attachments to the tag remain in place, so be sure the move is consistent with the constraints that have been defined for the destination tag group.
To move a tag to another group:
You will want to combine tags when you have two or more that have the same meaning. For example, let's say your data comes from a survey where respondents typed their Tenure into a free-style field and some entered "more than three years" while other entered "3 yrs". In such a case you will want the tag group to consist of tags that are a normalization of the values provided by the respondents.
When you combine tags, all content tagged by the consumed tag is tagged by the consuming tag unless it is already tagged by it.
To combine tags:
If you attach a tag to an entity by accident, or if for some other reason you no longer want to have a tag attached to it, you can detach it as follows:
Symphony's Chat Widget provides chat capabilities between you and other team members that have the same project open as you. You can optionally invite Symphony Support to join your chat.
The Chat Widget displays a list of all team members currently connected to the project you have open. When a team member makes an update to the project, their name briefly changes color and an icon indicates the view they were in when they made the change. When anyone sends a chat message, all members in the list receive it. If a connected team member doesn't have the chat open, a notice is displayed for them to let them know that chat is available.
To open/close the Chat Widget
To send a chat message, click the area to the left of the Send button then type your message. When you are ready to send it, either click the Send button or press the Enter key.
Cecil is our help bot. It is familiar with Symphony's help file so it is able to display topics or links to topics based on chat messages that you send.
To invite Cecil to chat, click his icon on the Chat Widget toolbar. Clicking again sends Cecil away.
When you minimize the chat window, it is removed from the screen but remains active. So if a team member sends a chat message, the window opens back up. This is in contrast with closing the window, where receipt of a chat message only displays a notice.
To minimize the chat window, click the minimize button on the chat window caption.
When Cecil returns a list of topics in response to your chat question, it includes a thumbs-down link. When you click this link, it lets Cecil know that the results it returned didn't answer your question. Cecil performs a deeper search to see whether there is anything else it can give you that might help.
When you view a topic that Cecil has returned, you have available a link that takes you to the same section in Symphony's help file. When you click this link, Symphony's help file is opened in another browser instance and you are placed at the relevant topic.
To have a chat transcript sent to your email account, click the icon on the Chat Widget's toolbar.
IMPORTANT: Many of Symphony's AI-related features rely in part on external providers. As such, the features are disabled for each project until you enable them. To ensure safe use of confidential data, please consider using Symphony's Maskifier widget on projects before you enable AI features.
This section covers various AI-related features available in Symphony.
Before you can use Symphony's AI features, you must enable them. This involves two steps.
Technically, Symphony's Maskifier performs no AI function directly, but it figures prominently in some of Symphony's AI features to provide safe and anonymous artificial intelligence. As such, it is important to be familiar with it before using Symphony's AI features.
The Maskifier enables you to hide entities such as names of people, companies, products, and so forth in your data prior to sending data to external resources (that is, outside Active Java's servers) that supplement or enhance Symphony's AI features. This is done automatically, so once you designate words for masking, you can forget about it.
The Maskifier also extends to reports, enabling you to have words redacted. This behavior is optional and is available for Feedback reports. You can also set it for individual queries on the Queries page.
Each word designated for masking is applied across all your projects, not just your current project.
The Maskifier is a widget located on the Widget's toolbar in the bottom-right of Symphony. Its icon is a theater mask. Clicking the mask toggles the widget open and closed.
The best place to work with the Maskifier is in either Review or Coding view. Both these views enable you to mask words in your coding structure and in comments. Soon, masking capabilities will be extended to People view.
The following instructions assume you are in Review or Coding view.
To reveal the Maskifier, click its icon in the lower-right corner of Symphony. Click the icon again or the widget's close button to close it.
The Maskifier provides two ways of working with masked words. They are accessed by the "Project" and "Global" buttons located at the top of the widget. Changes made with the Project button active are alway reflected in the Global list of masked words. Changes make with the Global button active are reflected in the Project list only if the word exists in the current project.
When the Project button is active:
Clustering is a powerful technique employed in the Symphony software for organizing and categorizing textual data, such as comments, into meaningful groups called clusters. This process involves grouping similar comments together based on their inherent characteristics, enabling users to gain valuable insights and extract patterns from large volumes of unstructured text. Symphony employs advanced algorithms to automatically identify common themes, topics, or sentiments within the comments, effectively assigning them to appropriate clusters. These clusters serve as labels, referred to as codes, and allow users to navigate, analyze, and manage the comments efficiently. By leveraging clustering in Symphony, users can uncover hidden patterns, identify emerging trends, and make informed decisions based on a comprehensive understanding of their textual data.
Symphony incorporates a set of algorithms that collectively could be called "Adaptive Silhouette K-Means++". Normally, the "K" in K-means is a user-defined parameter for specifying how many clusters the software should create. With Symphony, this is an optional setting. By default, Symphony samples your data using different values of K and different "seed" comments. The result that Symphony keeps is the one with the highest average "silhouette score". Symphony also produces code labels from the contained comments. You as the analyst then take charge of making final coding decisions and interpretation of the results. Essentially, your coding efforts begin after an initial coding pass has been performed.
Here are some useful details about Symphony's clustering:
Symphony's Sentiment feature helps you gain deeper insights into the opinions and attitudes expressed in a set of comments. By analyzing the language used in each comment, the feature determines whether the comment is positive, negative, mixed, or neutral. The sentiment is displayed with color-coded hearts, making it easy to understand the overall sentiment of each comment. Knowing the sentiment of each comment can be especially helpful when evaluating the quality of your code. For example, if your code reflects negative sentiment, most of the comments inside that code should also have negative sentiment. You may want to inspect comments with positive sentiment to see if they conflict with the intended tone.
The Sentiment column is an optional display in People and Coding view. You can include it in the display by adding the column using the column picker cog above the list where you want to include it.
Note that Sentiment is not calculated automatically. To calculate it or to update it after making significant changes to the underlying comments:
If you right-click a code, the sentiment is calculated for each comment in the code and it's subcodes. You can also select a range of comments and have them calculated at the same time.
You can also manually set sentiment:
Note that once comments are moved, the overall fit of other comments might be affected. This is because Symphony maintains statistics in real-time, recalculating where each comment fits best everytime one is moved.
Symphony's Autocoder automates the process of assigning codes to comments. After establishing a robust coding structure, leverage this feature to efficiently categorize newly imported content. Symphony narrows the scope of codes to match each comment based on its current location in the coding structure.
The comment is located in: | Symphony will consider these codes: |
The "Unassigned" code or one if its subcodes | The deepest codes in each branch of the coding structure |
A code that has subcodes | The subcodes of the code |
A code that does not have subcodes | The peers of the code |
In effect, Symphony narrows the scope -- if it can -- of codes that are contextually more appropriate. The assumption is that the comment is inside its current code for some reason; for example, it was imported as an answer to a question and the question is a code in your coding structure.
To initiate the Autocoder, you need to specify a scope of comments to be autocoded. The scope is based on what you currently have selected, and can be summed up as follows:
Selection: | Comments Autocoded: |
One or more comments | The selected comment(s) will be autocoded |
One or more codes | The comments directly inside the selected code(s) will be autocoded; comments in subcodes of the selection will not be included. |
One or more respondents in People view | The comments belonging to the selected person/people will be autocoded, regardless of where they are located in the coding structure. |
So, to summarize, a scope of comments is specified, and the candidate codes each comment can be autocoded to are the ones in the vicinity of the comment, unless the comment is in Unassigned, in which case all thematic codes are candidates.
Use this scenario when you have imported new data for specific respondents
Note that once comments are moved, the overall fit of other comments might be affected. This is because Symphony maintains statistics in real-time, recalculating where each comment fits best everytime one is moved.
Since comments are moved based on Symphony's analysis, understanding the changes might be challenging. To aid comprehension, Symphony indicates moved comments and their new codes with a small red circle next to the respective icons. These indicators remain visible until you modify/move an affected comment or run the autocoder again.
Given a comment that Symphony believes fits better in another code, this feature gives you an explanation as to why it fits better and provides a quick way to move it to the better code. When activated, this feature gives you:
So between the comment's similarity score, average silhouette score, and the narrative, you have a variety of ways for assessing whether to move the comment or leave it where it is.
Note that the results of each of these methodologies can point in different directions from what is expected. This is due in part to the fact that while the comment fits best in one code based on its Similarity to the underlying comments, it might actually cause the overall quality of either or both codes to degrade after being moved.
This feature is available only in Coding view with the Similarity column displaying, and can be performed as follows:
Symphony's Code Suggestion Component helps you code by creating a shortlist of potential codes for a comment. Symphony does this by analyzing the text of your codes, as well as the text of the comments in each code, and determines which ones best match the text of the comment. The effect of the comment being analyzed is removed so that it doesn't bias the results toward the way it is currently coded.
The Code Suggestion Component is available as an optional feature in most places where a coding structure is displayed. It contains the following visual attributes:
Symphony's Comment Suggestion Component helps you by listing comments that Symphony's AI identifies as fitting best in a highlighted code. Symphony analyzes all the comments that are not coded to the code you have highlighted, and selects the ones that score highest for that code.
The Comment Suggestion Component is optionally available in most places where a coding structure is displayed. It contains the following visual attributes:
Symphony's AI Code & Comment Rewrite Component assists you with creating the text for codes and comments that best communicates the important ideas in your project. This feature works in conjunction with an API published by OpenAI.
For Code Rewrites, Symphony analyzes several comments within the code to generate representative text. If the code is in the top tier of your project, the text resembles a high-level category. If the code is further down your coding structure, the text resembles a thematic statement that captures the primary findings or directives implied by the underlying comments.
For Comment Rewrites, Symphony makes the text more concise, clear, and professional while preserving the original meaning and intent. Typos and grammar are also checked.
This component can be found in the following places: next to the Content Editor in People and Coding views.
Initially, the component is hidden from view and can be exposed by clicking the chevron to the right of the text to be replaced.
To rewrite a code:
To rewrite a comment
This feature writes a summary of the comments in a code and its sub-codes. The result is saved to the target code's Notes field. You can use the summary in a few ways:
To generate a code summary:
This feature is currently in beta.
This feature enables you to converse with the AI about your project and topics related to qualitative analysis. You choose some or all of your project as a scope, then the AI answers your inquiries into the project data.
There is no limit to the length of the conversation. However, once the AI's context window size has been exceeded, old messages will be removed to make room for new ones.
Not sure which AI features to use and when? Given various scenarios of what you might want to achieve, this section gives you step-by-step instructions on how to achieve them with the assistance of Symphony AI features. Just choose from the list of scenarios below and follow the instructions.
Remember, to use the AI features, your project must have AI turned on in its properties.
For data privacy reasons, please consider using Symphony's Maskifier feature first. Doing this, you can ensure that no sensitive information can be related back to an individual or organization.
This section covers various topics that don't fit into main topics.
Some lists in the Symphony interface include optional columns. Optional display columns are available when a list has a cog on its header row.
Here's a list of possible columns you can display:
Symphony's Content Editor provides a quick way to make edits to the text of codes and comments. It is located in People view and Coding view, at the bottom of the page, and it displays the text of the currently-highlighted code or comment. It provides an alternative to opening the code or comment Properties form.
In the Content Editor you can:
To edit the text of a code or comment:
Symphony's Spawn Project feature lets you create a new project from the meta-data of another project. Meta-data consists of:
Basically, the spawned project includes everything from your existing project except for Respondents and other data specific to them.
The Spawn Project feature is useful when you are starting a new project and you have an old project that is similar. To spawn a project:
Symphony's Clone Project feature lets you make an identical copy of a project. Usually you would do this if:
To clone a project, simply right-click its name in the Project List and Choose Clone Project. The clone will appear in your project list.
Note that if what you really want to do is create a backup, it is better to download the file (click Download Symphony File in the main menu).
This features enables you to move or copy content from one project to another. You might want to do this in situations such as:
To move and/or copy content between projects:
When the move/copy is performed, Symphony ensures that all related data exists in the destination project. For example, if you copy comments, Symphony will create codes and respondents where equivalents don't already exist.
Let's say you generate a Feedback Report and while reading it you discover typos in some of the comments, or maybe you want to change the wording of some codes. If this is the last revision of the report that you'll be running and if you are finished with the project file, this is fine. If however you have more work to do, you will want the edits to be reflected in the project file so that the next time you run the report the edits will be carried over to it.
Symphony's Import Feedback Report Edits feature synchronizes report edits with Symphony content. Specifically, Symphony can apply text changes you make in a report to the corresponding codes and comments in your project file. Note that the imports are limited to text changes to existing items; new paragraphs are not added, and the order of codes and comments is not carried over.
To import report edits:
The Search Widget functions in People, Coding, and Review views, allowing you to find comments containing a specific word or phrase and optionally replace it.
Launch the Search Widget by doing any of the following:
Cecil is our AI Bot. In addition to lurking behind the scenes passively driving features such as Symphony's Code Suggestion component, Cecil can also help you learn what you need to know about Symphony. In this capacity you always interact with Cecil primarily through Symphony's Chat widget.
At the bottom of Symphony's screen, there is a field where you can ask a question. Pressing the Enter key opens Symphony's Chat widget with Cecil in attendance, who then provides a response to your question. From there, you can follow links to topics that Cecil found or down-vote what Cecil provided. In the latter case, Cecil will provide you with other alternatives.
Some lists allow you to customize which columns to display and in what order. Some of the available columns are built into Symphony, and some are driven by your data. For example, if you have demographics defined, they will be available as columns in the People tab.
Because the columns can be driven by your data, they are specific to each project. Therefore, your column settings are retained on a project-by-project basis. Each new project starts with the same default columns.
To choose which columns to display:
The columns are added in the order they are chosen. So if you want to change the order, you can de-select then re-select columns until they are in the order you desire.
Flags provide you a way of "marking" specific codes or comments. The meaning of the flag can be whatever you want it to be. For example:
To view or change your global settings, click the menu button followed by Preferences.
This section defines various concepts and how they are implemented in Symphony.
Demographics are custom attributes pertaining to respondents. Examples are Gender, Location, and Tenure. Each demographic has "demographic values" such as Male and Female for Gender.
The primary purpose of demographics in Symphony is to provide ways of grouping and/or filtering respondents to perform deep analysis or targeted reports. For example, with a Gender demographic you can compare the average comment sentiment for Males and Females. Or run a Feedback Report that contains comments belonging to respondents from a particular country.
Symphony implements demographics as "tag groups" and "tags". A demographic such as Gender would be a tag group, and Male and Female would be tags within the group. What differentiates a demographic from other tag groups is that you define the tag group to be used for respondents only.
An interview guide is used for transcribing interview notes. In addition to representing the results of an interview with an interviewee/respondent, it is the means by which interviews are usually imported into Symphony. Here are formatting guidelines:
Tags and Tag Groups are used to "mark" content. Tags can be used in many ways:
Having trouble? This section covers some of the more common issues.
The most likely cause is there has been a Symphony update and your browser has cached versions of files. Try to get your browser to download the updated files by doing the following:
Browsers are all different when it comes to caching files. Chrome is very aggressive with caching, and is known to be stubborn about refreshing files that it has updated recently.
The more time that goes by, the more likely your browser will be willing to refresh updated files.
This section provides a high-level outline for various types of projects. Consider each process step to be a major task that needs to be performed; the specifics should be aligned with the way your organization works and with the goal of your project.
This type of project usually seeks to answer questions pertaining to relatively small groups of people that have a shared knowledge or interest that brings them together. Examples are:
Each task does not necessarily need to be finished before continuing on to the next. Symphony fully supports an iterative approach to project completion. For example, you can add more demographics at any time.
Task | Details |
Create a new project. | For this type of project all you need is to give it a name at Start a new project. Leave the Multi-coding box unchecked. When you have comments that contain more than one theme, you should use the Split Comment feature. Coding the same comment to more than one code makes your verbatim reports confusing. |
Add information about the team members being assessed. |
As a minimum, this is the name of each interviewee and an optional title. It is often useful to include at least one demographic that might provide deeper insights into how or why interviewees are grouped around themes.
|
Add Comments from Interviews |
Use one of these methods to get comments into Symphony:
|
Code the Comments |
How you go about this is a function of how you approach qualitative analysis in general, and is also dependent on the amount and composition of data you have. Here's what Symphony has to assist you:
|
Perform post-coding analysis |
This is when you develop your narrative around your findings and is where your talent as an analyst prevails. To support you in your analysis, Symphony provides the following:
|
Create Feedback Report |
Guidance on how to create a template is available here. It is a best practice to create a general template that you can make a copy of and modify for each project. Run your report in the Feedback Report tab and distribute it for review. Reviewers' edits to codes and comments in the report can be imported back into Symphony using the Feedback Edits import option. |