Glossary of market research and Protobi terms

Welcome to the Protobi Glossary! This page serves as a helpful reference for the key terms and concepts used in Protobi, your go-to platform for streamlining research and data analysis. Whether you're new to Protobi or a seasoned user, this glossary will help clarify any unfamiliar terminology and ensure you're fully equipped to navigate our platform with ease.

Market research terms

• Survey
  A set of questions asked of respondents to help a researcher understand the market.

• Wave
  Surveys are often fielded in waves. Many surveys are one-and-done with a single wave. Other surveys, often called tracker surveys, are fielded in multiple waves over time. Waves may be annual, quarterly, monthly, weekly, or more frequent. Some surveys are fielded continuously, and waves are defined after the fact by grouping respondents into particular weeks or months.

• Respondent
  A person who answered the survey.

• Question
  A query in a survey that you, as the research designer, wish respondents to answer — such as "How old are you?" or "Which of the following products have you purchased in the last month?"

• Response data table
  Survey engines export data as data tables, typically in a format such as Excel, SPSS, or CSV (extensions .xlsx, .sav, .csv). Data tables typically have one row per respondent and one or more columns per question. A survey question may correspond to more than one column, especially if it is a compound or grid question. For example, in "Which of the following products have you purchased in the last month?", each product may have its own data column. Questions typically have question numbers to identify them, such as Q1 or Q2a.

• Frequency table
  Shows the frequency of each response. For instance, for a question like "How old are you?", a frequency table would show the number and percent of respondents at each age.

• Cross table
  A cross table (often called a crosstab) shows the correlation of two or more questions. For example, we may wish to see product interest crossed by age of respondent, to understand how product interest varies by age. The dependent variable — in this case, Product interest — is called the stub, and the independent variable — in this case, Age — is called the banner. Each stub value gets a row, and each banner value gets a column.

• Loop
  Surveys may include subsections that are asked multiple times of each respondent. For instance, a doctor may be asked about three recent patients, or a consumer may be asked about five recent shopping experiences. In the survey program, the repeated questions may be programmed as a loop in which they all share the same question ID, with another index tracking the iteration. This is efficient for studies that ask about multiple instances, and we recommend it for all patient chart studies and multiple-product surveys.

• Flat file
  A response data table that has one row per respondent, with many columns per respondent — flattening out grid questions and loop questions. You can recognize a flat file because it has many data columns, typically with variable suffixes separated by underscores (_) or dashes followed by letters/numbers. Also called a "wide" file.

• Stacked file
  A response data table that has one row per respondent per loop iteration. This is efficient because repeated questions keep the same identifiers, and multiple iterations are exported as different rows that can be easily aggregated. Also called a "tall" file.

Protobi terms

• Attributes
  The properties represented by "key":"value" pairs that you see in each element's JSON. Different attributes can take a string, numeric, array, or object as a value.

• Autogroup
  A feature of the advanced button that organizes elements into groups based on the name of the data column.

• Children Attribute
  An array (list) of child elements, separated by commas.

• Compact Group
  A group with child elements that have collapsed distributions. Each child element will show a compacted value (e.g., mean, top 2 box, etc.). The full distribution can be seen when the triangle next to a child element is pressed.

• Compactto Attribute
  A value or an array of values (e.g., mean, top 2 box, etc.) that are displayed when a group is in its compact (collapsed) form.

• Complement Testing
  Crosstab significance testing where each column is compared with the average of all other columns (excluding itself).

• DisplayKey
  A string alias assigned to a key for display purposes. If there is no displayKey assigned, an element will show the key label as the default.

• Element Attribute
  Seen as an option in the "More properties…" dialog. Reference a key in the element attribute to use a different question as the data source. Data is interpolated from that element, after applying its coding, transforms, etc.

• Elements
  Each question in your survey is represented as a chart in Protobi, often referred to as "Elements".

• Field Attribute
  Seen as an option in the "More properties…" dialog. Reference a key in the field attribute to use a different question as the data source. Unlike using the element attribute, using field will not apply the original element's coding, transforms, etc.

• Footnote
  Text for footnote at the bottom of an element.

• Format
  The label that is attached to a raw value.

• Frequencies
  Frequencies are the bars seen in a chart. They represent how often a value is chosen. Toggle the "#/%" on the toolbar to see frequencies as counts rather than percentages.

• Global Filter
  Filters saved as a named subset in the global button. When a global filter is on, it limits both current filters and baseline filters. Scenarios only limit current filters and are still compared to whatever population is set as the baseline (usually all respondents by default).

• JSON
  JavaScript Object Notation is a minimal, readable format for structuring data. JSON is the basis of Protobi elements, which determines how questions are displayed (structure, formats, etc.).

• Key
  The column name in the data file that directly correlates to a question.

• MaxValue
  Reference value by which to scale bars for this element. Useful when compacting to means, so that bars can be drawn for means relative to some value.

• NA Values
  A value that is blank in the data file. This can be due to a skip pattern, survey logic that doesn't force an answer, or a profile variable that doesn't have a recorded value.

• Parent
  A group with sub-elements (children). Parents often do not contain data values themselves.

• Project Elements
  Elements are the code that determines how each question is displayed in Protobi. Protobi uses JavaScript Object Notation (JSON) code. Project elements aggregate all the JSON elements in the project.

• Scenario
  A current filter, or a combination of current filters, saved as a named subset in the scenarios button. Scenarios only limit current filters and are still compared to whatever population is set as the baseline (usually all respondents by default). If you want to limit the survey to a particular subset and make that subset your baseline, use a global filter instead.

• Select an Element
  To select an element, press on the header of the element. This includes the display key, or the white space to the right of the display key. Another way to select an element is to press the context menu and choose the first option.

• ShowMissing Attribute
  Determines whether to show or hide missing values. "Yes" means always show, "No" means never show, and the default means using the current setting from the [NA] button in the toolbar.

• Split
  Delimiter to split values for this element. For example, "," will split a comma-delimited string into a multiple-response array.

• Toolbar
  The toolbar is the row of buttons seen at the top of the tool. Different buttons let you save, modify, and export the data.