PK ! Úsc§ § [Content_Types].xmlPK ! ˜Úë‹' ' _rels/.relsPK ! �ÎC C xl/workbook.xmlPK ! «!,n§ § xl/_rels/workbook.xml.relsPK ! `ÙAvZ Z
xl/styles.xmlPK ! 1¥Ë’+ ’+ xl/worksheets/sheet1.xml#StageCheckHow to do itStatusNotes / what I changed
1Before you startKeep an untouched copy of the raw dataSave raw file as read-only; work on a copy named with a date or version.Not started
2Before you startHave a codebook readyVariable names, labels, valid ranges, missing codes, reverse-coded items.Not started
3StructureOne row per respondent, one column per variableRemove merged cells, blank rows and notes inside the data table.Not started
4StructureClear, unique variable namesNo spaces or duplicates; match the codebook.Not started
5StructureCorrect data type for each columnNumbers stored as numbers, dates as dates; no text like "5 years" in numeric columns.Not started
6DuplicatesCheck duplicate IDs and duplicate rowsSort by ID or use Remove Duplicates; decide which record to keep and log it.Not started
7Missing dataCount missing values per variable and per respondentUse COUNTBLANK or a pivot table; note the pattern (random or systematic).Not started
8Missing dataMake sure missing codes are consistentReplace 99, -1 or "NA" with a single code or blank before analysis.Not started
9ErrorsCheck values outside the valid rangeFilter each variable; e.g. a 1-5 item containing 55 or 0, age of 250.Not started
10ErrorsCheck inconsistent answerse.g. age 15 but "retired"; end date before start date.Not started
11ErrorsCheck spelling variants in text categoriesStandardise "Lahore", "lahore ", "LHR" to one label (TRIM and PROPER help).Not started
12OutliersInspect extreme numeric valuesUse a boxplot or sort; check whether it is an entry error before treating it as an outlier.Not started
13RecodingRecode reverse-scored itemsNew value = (max + min) - old value, e.g. 6 - x on a 1-5 scale.Not started
14RecodingCreate derived variables and scale scoresKeep original items; add new columns for totals, means and groups.Not started
15QualityLook for straight-lining and implausibly fast responsesSame answer on every item; completion time far below the median.Not started
16FinalRe-run frequencies and range checksConfirm that every issue found is fixed and nothing new was introduced.Not started
17FinalWrite a cleaning logList each change: variable, rule, number of cells, date. Save with the data.Not started
|
Items doneCOUNTIF(E2:E18,"Done")
Items applicableCOUNTA(E2:E18)-COUNTIF(E2:E18,"Not applicable")
ProgressIF(D21=0,0,D20/D21)
|
Yellow cells are for you: pick a status from the drop-down and add notes. Source: analyticshorizons.com
"Not started,In progress,Done,Not applicable"PK ! Úsc§ § [Content_Types].xmlPK ! ˜Úë‹' ' Ø _rels/.relsPK ! �ÎC C ( xl/workbook.xmlPK ! «!,n§ § ˜ xl/_rels/workbook.xml.relsPK ! `ÙAvZ Z
w xl/styles.xmlPK ! 1¥Ë’+ ’+ ü xl/worksheets/sheet1.xmlPK € Ä: