The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Tanzania » Banking questions in Men's Questionnaire (DHS 2022) (Mis-routed questions in Men's Questionnaire on bank account and use of mobile phone to make mobile transactions)
Re: Banking questions in Men's Questionnaire (DHS 2022) [message #29244 is a reply to message #29201] Tue, 14 May 2024 13:31 Go to previous message
Janet-DHS is currently offline  Janet-DHS
Messages: 829
Registered: April 2022
Senior Member
Following is a response from DHS staff member, Tom Pullum:

Thank you for alerting us to this inconsistency between women and men in the questionnaire and tables (for the Tanzania 2022 survey) for using a mobile phone for banking transactions in the past 12 months. I will re-state what you said in terms of the numbers in the final report.

Both tables 15.6.1 (women) and 15.6.3 (men) have a footnote saying "Respondents were asked about use of a mobile phone for financial transactions whether or not they owned a mobile phone." However, I have checked, and this footnote only applied for women, not for men. I will paste here a table from TZMR82, in which I try to match the 21.7% in column 3 of table 15.6.3.

. tab mv169b mv169a [iweight=wt],m col

 
   last 12 |

months use |

    mobile |

telephone |

       for |

financial |     owns a mobile

transactio |       telephone

        ns |        no        yes |     Total

-----------+----------------------+----------

        no |  1,317.95  93.589638 |  1,411.54

           |     91.78       2.16 |     24.49

-----------+----------------------+----------

       yes | 118.07353  1,134.521 |1,252.5945

           |      8.22      26.22 |     21.74

-----------+----------------------+----------

         . |         0  3,098.865 | 3,098.865

           |      0.00      71.62 |     53.77

-----------+----------------------+----------

     Total | 1,436.024  4,326.976 |     5,763

           |    100.00     100.00 |    100.00
This table matches the n for men (5,673) but it confirms that there was a skip in the questionnaire for men, such that some men were NOT asked the banking question. The following table shows that men who did not have a SMART phone were not asked the banking question.

. tab mv169b mv169c [iweight=wt],m col
 

   last 12 |

months use |

    mobile |

telephone |

       for |

financial |  is respondent's mobile phone a

transactio |           smart phone

        ns |        no        yes          . |     Total

-----------+---------------------------------+----------

        no |         0  93.589638   1,317.95 |  1,411.54

           |      0.00       7.62      91.78 |     24.49

-----------+---------------------------------+----------

       yes |         0  1,134.521  118.07353 |1,252.5945

           |      0.00      92.38       8.22 |     21.74

-----------+---------------------------------+----------

         . | 3,098.865          0          0 | 3,098.865

           |    100.00       0.00       0.00 |     53.77

-----------+---------------------------------+----------

     Total | 3,098.865  1,228.111  1,436.024 |     5,763

           |    100.00     100.00     100.00 |    100.00 

Because this is a discrepancy in the questionnaire, I don't believe there is any way to modify table 15.6.3 to correspond with table 15.6.1. It WOULD be possible to modify table 15.6.1 to correspond with table 15.6.3, by re-classifying the women who did not have a smartphone as NA on the banking question (but including them in the denominator). I would not recommend that. As a general rule, NA cases should not be included in the denominator of a percentage. You could re-calculate the percentages for both women and men by removing those without a smartphone from the denominator. This would give consistency but could be misleading. I will not make a recommendation, but I hope other users will suggest something.

I have informed our data processing staff of this discrepancy and also the staff who work on questionnaire design.
 
Read Message
Read Message
Previous Topic: 2005 geocoding data
Goto Forum:
  


Current Time: Thu Sep 12 22:16:07 Coordinated Universal Time 2024