01 / COUNT WHAT IS IN THE FILE
The export reports zero nulls.
The cells tell another story.
A literal CSV-null check alone overlooks empty text and numeric response codes. Their exact semantics are not confirmed, so the analysis counts them without converting them into invented responses.
LITERAL CSV NULLS
0Cells parsed as null in this CSV export.NEGATIVE-CODED CELLS
395,994Numeric cells containing negative response codes.BLANK-STRING CELLS
14,793Fields containing an empty string rather than a CSV null.✳The audit in one minute: Zero literal nulls does not imply a fully observed questionnaire. Numeric codes (including −1, −2, −3, −5, −96, −97, −98) require their original response definitions.
LOOK BEYOND THE BLANK-CELL CHECKTHREE CHECKS · ORIGINAL VALUES PRESERVED
01 · THE EXPORT
0Literal null cells
A simple null counter would label the dataset complete.
02 · THE AUDIT
395,994Negative codes
The codes were inventoried—not silently treated as missing or valid.
03 · THE CONSEQUENCE
713Breakoff flags
Survey flow and case status need review before comparing groups.
03 / INTERPRET WHAT SURVIVES
A clean-looking export
still needs a codebook.
713 cases carry an RQI_breakoff flag; 710 have an empty Extraversion field. The counts show why missingness, survey routing, and measurement rules should be reviewed before interpreting personality comparisons.
WHY A CODEBOOK IS ESSENTIAL
Literal-null check0The parser detects no null-valued cells.
Code and blank audit14,793Additional blank-string cells, beyond the negative-code inventory.
Neither negative codes nor blanks are automatically evidence of missing data with a known cause. No undocumented imputations or score corrections were performed.
01 / EXPLORATORYZero nulls does not mean complete
What we found → The table has no literal CSV null cells, while 395,994 numeric cells carry one of the observed negative codes.
What to do → Audit coding conventions, skip patterns and blank text separately from ordinary NA checks.
02 / EXPLORATORYCompletion is a measurable issue
What we found → RQI_breakoff is 1 for 713 records, the same number as cases whose status differs from 6.
What to do → Reconcile process-quality metadata before choosing an analytic population.
03 / EXPLORATORYPersonality scores are not always populated
What we found → 710 Extraversion_mean entries are blank strings even though the CSV has zero nulls.
What to do → Convert blanks explicitly for score-specific analyses and report resulting sample sizes.
04 / EXPLORATORYAn apparent device difference is exploratory
What we found → The reported device-code group comparison needs a codebook to interpret groups 0 and 1 correctly.
What to do → Do not describe demographic or device effects until coding and sampling weights are verified.
DATA LIMITATIONS · UNRESOLVED- The exported CSV contains no literal nulls but uses blank strings and negative response codes extensively. Treating all these as valid survey answers is unsafe.
- Meaning of codes -1, -2, -3, -5, -96, -97, -98 is unverified without the survey codebook; negative-coded cells were counted, not silently recoded.
- A raking-weight variable exists, but its target population, weighting strata and intended inference are undocumented here. No population estimates claimed.
- Device-type codes and all personality scoring keys need the source instrument. The device group comparison is diagnostic only.
- The status and RQI_breakoff flags align numerically but the codebook must confirm their operational meaning.
The exact meanings of -1, -2, -3, -5, -96, -97, and -98 are not supplied. The published sample frame, raking calibration and codebook need verification.
