這一個階段在資料分析循環(Data Analysis Cycle)中,是必要,而且可以讓你節省許多時間,避免你進行資料分析或查核上的偏差。
資料分析與查核完整性(Integrity)的相關問題包含:
你所取得的資料都符合所申請的內容
在記錄當中的相關欄位計算皆產生正確的金額
數值型態欄位皆儲存數值型態的資料
每筆記錄皆屬惟一性,沒有重複的情形
日期欄位的資料皆是有效日期
儲在欄位內的資料皆符合預期認知
欄位間的關係符合一致性和邏輯性
依據ACL 105 Certified Training教材所提供的資料完整性驗證檢查表,參考如下:
To Check
Use
To Ensure
Validity
Verify
Data types and field definitions match.
Control Total
Count
Number of records matches control totals.
Total
Numeric field totals match control totals.
Statistics
Record and field totals match control totals.
Bounds Values
Statistics
Data falls within correct bounds
Between()
Data falls within correct bounds.
Missing Items
Gaps
Values are not missing from sequences.
ISBLANK()
Fields do not contain inappropriate blanks.
Duplicates
Duplicates
Tables do not contain duplicate records or inappropriate duplicate values.
Reliability
Computed fields
Calculations in source data are correct.
Reasonableness
Various commands
Data meets expectations.
Relationships
Various commands
Data is consistent across fields.
其餘後續再進行討論。
David
| 2 |
| |
|