Importing and Preparing Data in AddMaple
This article explains what happens when you load a dataset in AddMaple, how the app detects and tidies columns, and how to review or change those decisions.
Supported file types
- CSV
- Excel (
.xlsx
,.xls
) - SPSS (
.sav
) - Exports from most survey tools (e.g., Qualtrics, SurveyMonkey, Typeform, Google Forms)
No reformatting is required before upload.
What AddMaple does on upload
- Scans headers and a sample of values.
- Assigns a column type:
- Numeric (e.g., Likert 1–5, counts, percentages)
- Categorical (single-select)
- Multi-select (checkbox sets)
- Text / open-ended
- Date/Datetime
- Boolean (Yes/No, True/False, 1/0)
- Merges columns that belong to one question but were exported as many columns (common for single-selects and multi-selects).
- Groups related columns that share the same stem and scale (Likert/grid questions).
- Generates readable names for merged and grouped items by removing repeated stem/answer text often embedded in headers.
All steps are editable in Manage columns.
Merging columns (single-select and multi-select)
Many survey exports use one column per answer option. AddMaple collapses those into a single column so you analyze one variable instead of hunting across many.
Example: single-select exported as multiple dichotomies
Question: "What is your favorite color?"
Raw export (one column per option, with 1/0 flags):
- Respondent 1: Red=1, Yellow=0, Green=0, Blue=0
- Respondent 2: Red=0, Yellow=1, Green=0, Blue=0
- Respondent 3: Red=0, Yellow=0, Green=1, Blue=0
After merge in AddMaple (one categorical column):
- Respondent 1: Red
- Respondent 2: Yellow
- Respondent 3: Green
Example: multi-select (checkbox) exported as many columns
Question: "Select up to three colors you like."
Raw export (one column per option, 1/0 or Yes/No):
- Respondent 1: Red=1, Yellow=0, Green=1, Blue=0, Purple=0
- Respondent 2: Red=0, Yellow=1, Green=0, Blue=1, Purple=0
- Respondent 3: Red=1, Yellow=1, Green=0, Blue=0, Purple=1
After merge in AddMaple (one multi-select column):
- Respondent 1: Red, Green
- Respondent 2: Yellow, Blue
- Respondent 3: Red, Yellow, Purple
Notes:
- Common encodings (
1/0
,Y/N
,TRUE/FALSE
) are recognized case-insensitively. - If some options are empty in your file (no respondents chose them), they won't affect the merge; you can still keep or hide those options in labels.
Grouping related columns (grids / Likert scales)
Grid questions are usually exported as one column per item with the same response scale. AddMaple keeps the columns separate but groups them so you can view and compare them together.
Example: product attribute ratings (Likert 1–5)
Raw export:
- Respondent 1: Ease of use=5, Value for money=4, Customer support=3
- Respondent 2: Ease of use=4, Value for money=5, Customer support=4
- Respondent 3: Ease of use=3, Value for money=3, Customer support=2
After grouping in AddMaple:
- Group title: "Please rate the following aspects of the product"
- Group members (individual columns): Ease of use, Value for money, Customer support
In dashboards, the three items appear together. You can:
- view them side-by-side,
- filter by any item's value,
- open any single item to analyze relationships on its own.
Merge vs. Group: what's the difference?
Merge
- What it does: Combines many source columns that represent one question into one column (categorical or multi-select)
- Where you see the effect: You work with a single variable everywhere (pivots, filters, charts)
Group
- What it does: Leaves columns separate but links them as a logical set (common stem/scale)
- Where you see the effect: In dashboards, grouped items appear together; you can still analyze each column individually
Rule of thumb:
If you see ten checkbox columns for one question, that's a merge case.
If you see several attributes rated on the same scale, that's a group case.
Managing columns (review, fix, customize)
Open Manage columns from the dataset toolbar to inspect and adjust what AddMaple did.
You'll see
- The original columns from your file.
- Any merged columns AddMaple created, with their member columns listed.
- Any groups, with their member columns.
- Detected types (numeric, categorical, text, date, boolean).
You can
- Unmerge / re-merge: split a merged column back to sources, or select sources and merge them.
- Group / ungroup: add or remove columns from a logical group.
- Rename:
- Group title (e.g., shorten a long stem).
- Merged column name (e.g., remove redundant prefixes).
- Category labels (e.g., change
1
→Strongly disagree
).
- Change type:
- Convert to Date/Datetime if a date was detected as text.
- Convert numeric codes to categorical labels (or the reverse).
- Set Boolean when headers contain Yes/No flags.
- Reorder categories (e.g., Likert from Strongly disagree → Strongly agree).
Changes apply immediately to charts, filters, and pivots.
How titles are generated
Survey exports often repeat the question stem and the answer label in headers (e.g., Q12_Product: Ease of use
, Q12_Product: Value for money
). AddMaple:
- extracts the common stem for the group title,
- uses the varying part for item names,
- and, for merges, derives a concise merged column name.
You can override any of these in Manage columns.
Changing a column type (examples)
- Fix a date column
Detected as text? Change type to Date. If formats are mixed (e.g.,MM/DD/YYYY
andDD/MM/YYYY
), standardize them in the source or split the column before upload. - Code numeric Likert to labels
Values1–5
? Change type to Categorical and set labels:1=Strongly disagree … 5=Strongly agree
. - Treat 1/0 flags as Boolean
Change type to Boolean to simplify filters and summaries.
Common questions
The tool merged columns I don't want merged.
Open Manage columns → select the merged column → Unmerge. Then merge only the columns that belong together.
A multi-select wasn't merged.
Select the related checkbox columns in Manage columns → Merge. Ensure the headers correspond to the same question.
The group title is too long.
Rename the group in Manage columns. Item names remain unchanged.
My dates look wrong.
Change the column type to Date. If the source mixes formats, standardize in the source file to avoid ambiguity.