Introduction

You can import data into EyeOpenR insight in various file formats or directly from your EyeQuestion Environment. Below you can find a description of the various methods through which you can import your data into EO Insight.

How to import data directly from an EyeQuestion environment

In order for EyeOpenR Insight to import data directly from your EyeQuestion environment, you first need to have this function enabled both within your EyeQuestion environment and EyeOpenR Insight user account. To enable direct importing from your EyeQuestion environment or add this function to your EO Insight account please contact support@eyequestion.nl.

If both your EyeQuestion environment & EyeOpenR Insight user account are enabled for direct EyeQuestion imports you can follow the steps below to quickly import your data to EO Insight directly from your EyeQuestion Projects.

Step 1: Click the ‘Import Data’ icon in the sidebar to reveal the ‘Import Data’ sidebar or choose the ‘Upload from EyeQuestion’ option on the ‘Welcome’ popup that is displayed when you start the app.

Step 2: Click on the ‘Import EyeQuestion Project’ title to reveal the options for importing from your EyeQuestion environment

Step 3: Choose whether to anonymise the data retrieved from EyeQuestion (If anonymise is set to ‘Yes’ this will anonymise the Assessor Name and Login)

Step 4: Enter the Project Author. Only projects that are authored by the name provided here will be listed for retrieval [the author of a project can be found & adjusted on the ‘Summary’ tab within your EyeQuestion project]

Step 5: Adjust the period in which projects should be listed for retrieval. Projects will be listed if they have been modified from the current date until the number of months defined by this variable prior to the current date e.g. if this value is set to 6 a list will be generated consisting of all projects modified in the past 6 months and authored by the name provided in the author field

Step 6: Click the ‘Import EyeQuestion Project Data’ button

Step 7: Enter the secret API key for your EyeQuestion environment (if you do not know this ask your ‘EyeQuestion Environment Primary User’ or if you are a ‘Primary User’ contact support@eyequestion.nl), then click OK.

Step 8: In the ‘EyeQuestion Project Selector’ window select the row containing the project you wish to import, then click ‘Retrieve Project Data’ and your data will be imported to EO Insight

Importing data from an Excel (.xlsx) or CSV (.csv) file

EyeOpenR insight allows the importing of a wide variety of data structures. The attached datasets provide examples of file structures that can be imported to EyeOpenR Insight. The demo datasets aim to demonstrate two ends of the spectrum from a file containing multiple metadata sheets to a simple csv file.

For both xlsx files and csv files in order to be compatible with EO Insight the file must contain at least one table of data with column headings in the first row, each subsequent row is assumed to contain observations relating to individual judge - product combinations.

NOTE: Question types that code multiple products per row are therefore currently not compatible with EO Insight analysis methods (e.g. a subset of discrimination task question types, temporal method question types & conjoint question types), although this will be addressed in future versions EO Insight. 

If your file contains multiple sheets a sheet titled ‘Data’ or in its absence the first sheet is assumed to contain the raw data. When using the import wizard you will be able to examine and adjust the import process at various stages (see below sections for further details). 

Steps for importing your dataset from an .xlsx or .csv file into EO Insight

Step 1: Either choose the “Upload External Dataset” option from the ‘Welcome popup’ that is displayed when starting the application, or click the ‘Import Data’ icon in the EO Insight sidebar to reveal the ‘Import Data’ sidebar

Step 2: Select from the dropdown the origin/data structure of the file you wish to import (e.g. was your data exported directly from EyeQuestion or Compusense). This is optional as EO Insight will attempt to deduce the structure from the file uploaded.

Step 3: Click the ‘Upload’ button and then select the file containing the data you wish to import

Step 4: You will now see the ‘Quick Import Wizard’. This allows you to view and adjust the imported data (for further information see sections below). We recommend checking the data in the ‘Data’ table and the ‘Information’ tab before accepting the import.

Step 5: If you are happy with the data as displayed in the ‘Quick Import Wizard’ click ‘Import’ and your data will be imported to use in EO Insight

Quick Import Wizard

When you import your data, EO Insight will attempt to map your data onto the EO Insight data structure to ensure compatibility with the various analysis methods offered within the application. In both the ‘Quick’ & ‘Advanced’ import wizards you can then examine and adjust the resulting data structure before the data is used in EO Insight. The ‘Quick Import Wizard’ in EO Insight allows you to adjust key properties of the structure inferred by EO Insight. On the left side of the ‘Data' tab you will find drop down buttons where you can reassign a key variable to another column in your dataset or exclude columns from the import [see (A) in the above screenshot]. For example, if the import wizard failed to identify the variable coding your product ID you can click on the drop down ‘Select Alternative Product Column’ and then select the correct column from the list. The corresponding column will then be coloured to represent this change.

Advanced Import Wizard

The ‘Advanced Import Wizard’ in EO Insight allows you to adjust the metadata structure inferred by EO Insight and/or enhance your imported dataset with additional metadata. When you import your data EO Insight will attempt to map your data onto the EO Insight data structure (for more information see the section below: “Additional information on importing data and the EO Insight Data Structure”) you can then examine and adjust the resulting data structure in the wizard before the data is used in EO Insight. To access this ‘Advanced Import Wizard’ click the ‘Advance Import Wizard’ button in the bottom left corner of the ‘Quick Import Wizard’.

At the top of the ‘Advanced Import Wizard’ you will find the following tabs [see (A) in below screenshot]:

• Data: the Raw data imported into the EO Insight ‘Data’ table with EyeOpenR reference values displayed for Attributes (Q*), Products (P*), Judges (J*) & Category Labels.

• Attributes: the attribute / question metadata extracted from your imported file and mapped onto the EO Insight ‘Attribute’ table

• Products: the products / sample metadata extracted from your imported file and mapped onto the EO Insight ‘Product’ table

• Assessors: the panelist / judge / assessor metadata extracted from your imported file and mapped onto the EO Insight ‘Assessors’ table

• Sessions: the session / section metadata extracted from your imported file and mapped onto the EO Insight ‘Sessions’ table

• Projects: the project metadata extracted from your imported file and mapped onto the EO Insight ‘Projects’ table

• Information: Important information about any adjustments made by EO Insight when importing your data including whether any incompatible data was excluded

On the left side of each tab (other than the Data tab) you will find drop down buttons where you can reassign a key variable to another column in your dataset [see (B) in below screenshot]. The corresponding column will be coloured to represent this change.

If you wish to edit a cell value in any of the metadata tables (i.e. Attributes, Products, Assessors, Sessions, Projects) simply double click in the cell to access the edit function. Then edit the cell before pressing enter/return to save your changes. The cell should then be updated with your edited value. This can be particularly useful should you need to change the ‘Attribute Name’, ‘Datatype’, ‘Min’, ‘Max’ or ‘Parent_Question’ values automatically assigned to an attribute.

If you wish to import additional metadata you can upload a replacement metadata table by using the ‘Upload replacement table’ function present on the left side of each metadata tab [see (C) in below screenshot].

If you wish to download or adjust any of the metadata tables you can click the ‘Download Table’ button on any of the metadata tabs to download the table in excel format [see (D) in below screenshot]. You can then make any adjustments to the file in excel before using the ‘Upload replacement table’ function to replace the metadata table shown in the wizard.

How to import Panelist and/or Product metadata for analysis with EO Insight

The ‘Advanced Import Wizard’ also allows you to import variables from your metadata sheets into your Data table to be used for analysis and other EO Insight functions. For example if the age or nationality of your panelists are stored in your ‘Assessor’ metadata and you wish to use this for analysis or to filter your data. 

On both the ‘Products’ and ‘Assessors’ tab in the ‘Advanced Import Wizard’ you can find the ‘Select columns to include as additional variables for analysis’ option as shown below. Clicking on the drop down will reveal a list of all columns in the given metadata table. Select the columns you wish to be imported as new variables in your Data sheet, once selected they will appear as colour coded in the corresponding ‘Data Import Wizard’ table. Once you complete the import process these new variables will be present in your Data table with values mapped according to the ‘Product’ or ‘Judge’ columns in each table.

Additional information on importing data and the EO Insight Data Structure

Although it is not necessary to have this structure in your imported file, on importing your data to EO Insight it will construct a Data table containing your raw data, the first 5 columns of this Data table are always structured as follows in EO Insight:

• Judge - the Judge name/code/reference value

• Product - the product name/code/reference value

• Session - the session number or reference value

• Replica - the number of the replica or its reference value

• Sequence - the order or position of presentation for a given sample/product

When importing a dataset EO Insight it will attempt to identify the following Column Headings in your raw data table to map on to the Judge, Product, Session, Replica & Sequence variables that form the first 5 columns of the EO Insight Data table. If it is unable to find these column headings then it will attempt to deduce their values or assign default values. Below are the column headings it looks for:

• Judge: assessor, assessors, panellist, panelist, panellists, panelists, judge, judges, panelist_username, subject_id, consumer

• Product: product, sample, products, samples, sample_number, product_id

• Session: session, section, sessions, sections, section_name, section_number

• Replica: replica, replicas, rep_name, rep_number, rep, repetition, replicate

• Sequence: sequence, sample_position, sequences, rank

For files containing multiple sheets, EO Insight uses the following sheet names (irrespective of case i.e. upper or lowercase) to identify the corresponding data:

• Raw Data: data, dataset 

• Attribute / Question Metadata: attribute, attributes, question, questions

• Assessor / Panelist / Judge Metadata: assessors, assessor, panellist, panelist, panellists, panelists, judge, judges

• Product / Sample Metadata: product, products

• Session / Section Metadata: session, sessions

• Project Metadata: project, projects

Similar to the Data table there are a key set of columns on each metadata table within the EO Insight data structure. 

• Attributes Metadata Table

• Attribute: Attribute ID or reference number

• Datatype: the datatype of the variable (see below for explanation)

• Min: minimum  value for attribute / question

• Max: maximum value for attribute / question

• Display: Name of the attribute / question

• Label*: Labels for the categories / levels within the attribute / question (e.g. Label2 provides the name of the category assigned to any value of 2 recorded in the Data table for the given attribute).

• Assessors Metadata Table

• Judge: Assessor / Panelist / Judge reference number or ID

• Name: Assessor / Panelist / Judge name

• Products Metadata Table

• Product: Product ID or reference number

• Name: Product Name

EO Insight will attempt to identify these within your imported dataset. If they are not present it will attempt to deduce their values or assign default values. If you include metadata sheets it searches for the above column headings to map on to its key metadata columns.

The Attributes table assigns a ‘Datatype’ to each variable which dictates the given variables compatibility with the analysis methods/widgets available in EO Insight. Below is a list of the ‘Datatypes’ assigned by EO Insight:

• Nominal: Categorically discrete data such as name of a country, gender etc. 

• Ordinal: Quantities that have a natural ordering, e.g. the order of items placed in a line or the ranking of favorite foods. With ordinal data you cannot state with certainty whether the intervals between each value are equal.

• Interval: Similar to ordinal but the intervals between each value are equally split. e.g. data coming from line scales, temperature.

• Text: Textual data, e.g. remarks.

• JAR: Just about right data

• Binary: Can be either 1 or 0. Normally this type of data is used in pick any (CATA) data, discrimination test (Triangle, Tetrad etc.) and for true/false questions.

Once the data is imported you can use the ‘Data Table’ widget to inspect how your data has been imported into the EO Insight data structure. If you wish to download, edit and reimport your dataset; or simply want to see an example of how EO Insight structures data for use in the application in an Excel file you can do this via the ‘Export Current Dataset’ function in the ‘Export’ sidebar.

FAQs

How do I adjust my dataset after importing?

If you have imported your data to EyeOpenR Insight and then decide you wish to make a number of changes to your dataset you can use the ‘Download Current Dataset’ function in the ‘Export’ sidebar to quickly download the dataset that is currently active in your EyeOpenR Insight session. This will generate an excel file in the EyeOpenR data structure containing both Raw data and metadata sheets. You can then easily edit this file and then reupload into EyeOpenR once all your changes have been made.

How to change my attribute names?

On importing your dataset to EyeOpenR Insight it is possible some of your question/attribute names are converted to lowercase or adjusted in other ways as part of the import mechanism (future versions of EyeOpenR Insight will minimize such events). You can change the names of your ‘Attributes’ within the ‘Advanced Import Wizard’ by clicking in the corresponding cell in the ‘Display’ column on the ‘Attribute’ Tab and adjusting as required. Alternatively if you have already imported your data you can use the ‘Download Current Dataset’ function in the ‘Export’ sidebar to quickly download the dataset that is currently active in your EyeOpenR Insight session. This will generate an excel file in the EyeOpenR data structure containing both Raw data and metadata sheets. You can then easily edit this file and then reupload into EyeOpenR once all your changes have been made.

How do I define labels for category variables?

Should your dataset contain category variables (e.g. Nominal, Ordinal, JAR) it is possible these are coded as numeric even though they correspond to labelled categories. Within EyeOpenR Insight it is possible to define the label for each category within each category variable. This information is stored in the ‘Labels’ columns within the ‘Attributes’ metadata sheet (e.g. using the ‘Data Table’ widget with the metadata sheet view enabled, see “General Purpose Widgets” Knowledge Base article). On importing your dataset EyeOpenR Insight will attempt to identify category variables and their corresponding category labels. You can inspect these labels assigned during the import process by viewing the ‘Attributes’ tab using the ‘Advanced Import Wizard‘. To adjust any of the labels in the ‘Label’ columns simply click the cell you wish to adjust and then enter the new value. Should you have already imported your data use the ‘Download Current Dataset’ function in the ‘Export’ sidebar to quickly download the dataset that is currently active in your EyeOpenR Insight session. This will generate an excel file in the EyeOpenR data structure containing both Raw data and metadata sheets. You can then edit or add new ‘Label’ columns on the ‘Attributes’ metadata sheet to define category labels. Column headings must follow the logic ‘Label*’ where * is the numeric value coded for this category in the corresponding attribute column on the ‘Data’ sheet. For ‘Binary’ variables imported via the ‘External’ dataset import wizard values of ‘0' will automatically be assigned the label ‘Not Selected’ while values of ‘1’ will automatically be assigned the label 'Selected’.

What defines which variables are compatible with which widgets?

The ‘Datatype’ assigned to a variable dictates which variables are compatible with which widgets. The Datatype assigned to a variable can be viewed via the ‘Attributes’ metadata sheet (e.g. using the ‘Data Table’ widget with the metadata sheet view enabled, see “General Purpose Widgets” Knowledge Base article). During the import process you can define the ‘Datatype’ of each Attribute in your dataset on the ‘Attributes’ tab within the ‘Advanced Data Import Wizard’. Should you wish to know which widgets are compatible with which Datatypes then you can ask the AI Chatbot ‘General Assistant’ or view the widgets corresponding Knowledge Base article.

Why isn’t the data from my Temporal or Discrimination project loading?

Unfortunately, this early version of EyeOpenR Insight does not support all data types generated by all question types, for example currently EOi does not support all data generated by Discrimination and Temporal question types. This functionality will be available in future versions of EyeOpenR Insight but until this time please use EyeOpenR for analyses of such datasets.