Numbers, units and identifiers

Results are kept exactly as typed or imported and read as numbers when you calculate, using the study's numeric convention. Units and IDs are used as written.

Choose the numeric convention first#

Each study has one numeric convention. Set Numeric convention in Calculation settings on Set up before entering results. A study made from a file takes the file's convention.

The Data stage shows Decimal point or Decimal comma beside the unit.

How values are read#

A value written in the other convention is flagged. Correct it, or change the study's convention if the whole file uses the other one.

Thousands separators#

• Put exactly three digits after each thousands separator, as in 12,345,678 or 12.345.678. In a decimal-point study, 1,2345 is flagged.
• Remove spaces inside numbers. 1 234 is flagged.
• In a comma-delimited CSV, quote values that contain a comma, such as "1,234.5". See CSV and TSV files.

Scientific notation#

1.2e2, 1.2E2 and 5e-3 are accepted. With a decimal comma, write 2,5E1.

Signs and negative results#

A leading +, - or − is accepted. Negative results are flagged unless you tick Permit negative results in Calculation settings. It is ticked for new detection-limits studies.

Supported range#

These values are flagged:

• an absolute value above 10¹⁰⁰
• a nonzero value too small to calculate with, such as 1e-400
• more than 15 significant digits, such as 10000000000000001. Round it. Numeric Excel cells are exempt.

NaN and inf count as text. Dilution factors run from 1 to 10⁹.

Changing the convention after rows exist#

Once a study has rows, saving a new Numeric convention opens Change the number format?, because a value such as 1,234 can change meaning. Choose Change number format, then calculate again.

After a change, the results warn about imported rows that came in under the old convention. Check those rows. Rows typed in the Data stage, including dilution factors and volumes, get no warning, so check them too before calculating.

Units#

A study has one unit, set on Set up. Values are never converted between units.

• A unit in a result heading, such as Result (mmol/L), fills in a new study's unit. Imported into an existing study, it must match that study's unit. Case and spaces don't matter.
• A Unit column must match exactly, including case, so mg/dl and mg/dL differ. Rows in other units form separate groups. See Files that mix analytes, matrices, instruments or units.
• A limit must use the unit of the statistic it limits: the study unit for concentrations, % for percentages and ratio for a slope. The limit dialogs fill this in for you; a limit in any other unit stays Undecided. See Acceptance limits.

Interference and stability also set units on Set up for interferent concentration and time.

Identifiers#

Each study type has one ID column, such as Specimen ID, Material ID, Subject ID or Pool ID. Use coded IDs instead of patient names or record numbers.

• IDs are kept as written, so 00001 stays 00001. Excel drops leading zeros from numeric cells, so format ID columns as Text.
• Spaces at the start and end are removed.
• s01 and S01 are different specimens.
• IDs hold up to 80 characters.

Rows that share an ID are replicates of one specimen or material. Specimen and subject counts use distinct IDs. For example:

• In reference intervals, each subject has one row per partition. Repeat draws count once.
• In qualitative agreement, a specimen's first included row is its result.
• In a lot comparison, an ID starting with QC marks quality-control material.

Level, day, run and replicate#

These are labels. 1, L1 and Low all work. Spell each one the same way throughout, because Day 1 and 1 are different days.

A blank cell in a used column is flagged where the study needs a value. Level holds up to 80 characters, and day, run and replicate up to 40. Multi-day precision needs Day, Run and Replicate on every row. Linearity groups by expected concentration when the file has no Level column.

Other text fields#

Lot, role, storage condition, diluent and exclusion reason are text. Some must match a fixed list or a value set on Set up, such as role blank or low in detection limits. See the column reference. Case doesn't matter for role, category and storage-condition labels. Lot labels must match exactly, including case.