Data Quality(Data Cleaning)

Inconsistent responses are completely excluded.

Is it possible to have survey panel in Southeast Asia?
Are you worried about the quality of their answers?

With W & S Thailand and Vietnam survey panels, answer quality is our main priority.
Not only applying the screen control system, we also conduct visual review of the response, to ensure high quality for the research results and guarantee its validity.

In order to deliver the accurate data, the data retrieved by W&S will be checked through investigation as showed in the "logic check sheet". Any discrepancy within respondents' responses will be excluded. Thus we only deliver the valid and reliable data.

Data Cleaning Process

1.Before delivering the final valid data, we will present the Logic check sheet, where we specify the conditions that lead to incorrect or inconsistent answers.

2. Based on the conditions in Logic check sheet, we will perform the Data Cleaning process.
In case respondent' answers are alike/confusing, we will directly contact that particular respondent to get their clear confirmation.

Logic Check Sheet

Before processing and analyzing data, we will check the answers carefully based on the Logic check sheet to make sure the reliability of raw data.

Data cleaning - Case study

Case 1. Discrepancy between selected options of two questions (Illogical/Inconsistent Answers)

Recheck the following case:

If A_10 (Impressive Advertisement) / A_11 (Advertisement appear frequently) / A_17 (Promotion Campaign) in Q2 and A_7 (I have never known the advertisement and promotion campaign of this brand) in Q1 are selected at the same time.

Q1

Q2

Case 2. Discrepancy between future consumption intention within two questions (Inconsistent Answers)

Recheck the following case:

If A_1 (Very Likely) / A_2 (Somewhat Likely) of one brand in Q1 is chosen, but A_5 (Definitely not interested) of that brand in Q2 is also chosen.

Q1

Q2

Case 3. Discrepancy between Current Age and Highest Education Level (Inconsistent Answers)

Recheck the following case:

Recheck if A_2 (20 - 24 years old) in Q1 and A_7 (Doctor Degree/PhD) in Q2 are chosen at the same time.

Q1

Q2

Case 4. Speeders and Straight-Liners

Recheck the following case:

Recheck Matrix.
Recheck respondents who simply choose the same codes in each options and in a very short time (speeders/straight-liners)

Case 5. Improper Words/Spam

Recheck the following case:

Check text in A_11 (Others). If the text is improperly worded (irrelevant to the question asked or spam >> delete their result)

Case 6. Discrepancy between Device to find current job and Information about current job

Recheck the following case:

Recheck if A_7 (None of the above devices in Q1 and A_3 (Internet Jobs site), A_12 (Apply through website of recruiting company), A_13 (Social network sites) in Q2 are chosen at the same time.

Q1

Q2

number of panellist

15-Dec-2017 130,244

Free Direct HaiSurvey

Hotlines

Bangkok
+66 2 6530411
Jakarta
+62 29 022227
Ho Chi Minh
+84 8 3848 3731
Tokyo
+81 3 54217925
 

Company

Recruit

Contact

  • Bangkok +66(2)6530-411
  • Tokyo +81(3)5421-7925
  • Jakarta +62(21)2902-2227
  • Osaka +81(6)6390-1441
  • Hochiminh +84(8)3848-3731
  • Okayama +81(86)293-2934

Company

Recruit

Contact

  • Tokyo +81(3)5421-7925
  • Okayama +81(86)293-2934
  • Osaka +81(6)6390-1441
  • Jakarta +62(21)2902-2227
  • Bangkok +66(2)6530-411
  • Hochiminh +84(8)3848-3731