Table 2.
Specific detectable errors implemented in cluster sample survey.
| # | Class | Error definition | Number of errors | Error rate, % |
| 1 | Intentional redundancy | Gave different answers for the question (“Was your most recent birth in a health facility?”) in different sections of the questionnaire | 19/618 | 3.1 |
| 2 | Intentional redundancy | Gave different answers for the question (“Have you ever given birth?”) in different sections of the questionnaire | 10/961 | 1.0 |
| 3 | Intentional redundancy | Gave different answers for the question (“What was the date of birth of your most recently birthed child?”) in different sections of the questionnaire | 84/618 | 13.6 |
| 4 | Intentional redundancy | Gave different answers for the question (“Is your most recently birthed child still alive?”) in different sections of the questionnaire | 10/618 | 1.6 |
| 5 | Illogical response combinations: single question | Question is “Where you go to get medical advice or treatment?”; answer options included (“refused to respond” OR “unknown”) AND (“clinic” OR “drugstore” OR “community health worker” OR “traditional healer” OR “other”) | 2/895 | 0.2 |
| 6 | Illogical response combinations: single question | Question is “What are the signs of someone who can have ebola?”; answer options included (“refused to respond” OR “unknown”) AND (“fever” OR “muscle pains” OR “vomiting” OR “sore throat” OR “diarrhea” OR “bleeding” OR “other”) | 0/895 | 0.0 |
| 7 | Removal of “required” constraint | A required question (“Can people get Ebola from touching an Ebola patient?”) was skipped | 0/895 | 0.0 |
| 8 | Removal of “required” constraint | A required question (“Can people get Ebola from the air?”) was skipped | 0/895 | 0.0 |
| 9 | Removal of “required” constraint | A required question (“Can people get Ebola by touching or washing a dead body?”) was skipped | 0/895 | 0.0 |
| 10 | Illogical response combinations: multiple questions | Answers for a multiple-response question (“From whom did the child get treatment [for fever or cough]?”) were given; an answer was given to the following question (“From whom did the child get treatment FIRST?”) that was not selected in the previous list of responses | 0/325 | 0.0 |
| 11 | Illogical response combinations: multiple questions | Answers for a multiple-response question (“From whom did the child get treatment [for diarrhea]?”) were given; an answer was given to the following question (“From whom did the child get treatment FIRST?”) that was not selected in the previous list of responses | 0/202 | 0.0 |
| Total | 125/7817 | 1.60 |