The data behind the ballots

Tuesday November 14 2017

By JOHN WALUBENGO
More by this Author

Two weeks ago, the Independent Electoral and Boundaries Commission (IEBC) in possibly one of its most transparent moves, released a huge trove of data behind the voting in the repeat Presidential elections of October 26, 2017.

This is the data that was automatically recorded in the logs of the Kenya Integrated Election Management System (KIEMS) kit, as voters get identified before being allowed to vote.

Indeed the dataset is huge, given that it covers around 40,000 polling stations and runs to over a thousand pages.

Worse still, it is in ‘PDF’ format, which technically is not conducive for deep, incisive, interrogative analytics. Luckily, I got some emerging data scientists from the university to convert the data into a conducive format for analytics.

They managed to mine some interesting analysis that is worth sharing.

The first perspective shows the percentage of voters identified electronically through their fingerprints across each county [CLICK IMAGE].

Mombasa County had the largest percentage of voters identified biometrically (88 per cent) while Nyamira County had the lowest percentage number at 57 per cent.

Put differently, 12 per cent of voters in Mombasa County and 43 per cent of the voters in Nyamira County were allowed to vote using ‘other means’ that were not biometric.

Nyamira County therefore had the largest percentage of voters NOT identified biometrically, followed closely by Kericho at (40 per cent), Wajir at (38 per cent), Nyeri at (38 per cent) and Tharaka Nithi at (38 per cent) respectively.

UNREADABLE FINGERPRINTS

One can only hope that IEBC will be able to investigate and explain to Kenyans why a standard electronic voter identification kit (EVID) would behave so different across these different counties.

But which are these other methods that IEBC uses to identify voters?

During the contentious debates around whether to use complementary methods or not, it was argued that some voters may have unreadable fingerprints, perhaps arising from the nature of their work or accidents prior to voting day.

The election law was subsequently amended and IEBC allowed to clear voters outside the biometric identification framework.?

One way to clear voters non-biometrically is the ‘Alphanumeric’ approach. In this approach, the IEBC clerk would receive the voter's national ID card and type the corresponding ID number into the EVID Kit.?

If the voter is in the register, their profile would come up on the screen, and they would be cleared to vote.

In previous reports, IEBC confirmed that about 1.6million voters were identified through this alphanumeric approach. This is also in agreement with the released datasets.

The second category of non-biometric voter identification was for those voters whose biometric features were originally not captured. Such voters would be cleared under the ‘Document Search’ category. Approximately 400,000 voters were cleared through the ‘Document Search’ approach during the repeat elections.

Obviously non-biometric clearance has its weaknesses in that it can be abused by having absentee voters cleared.

SUPPORTING DOCUMENT

So there were additional procedural requirements to be completed upon clearing voters using non-biometric means, one of which is to make sure that for each voter cleared through the Document Search route, a form 32A needed to be completed and signed by agents as one of the supporting documents.

In terms of absolute figures, we are looking at around two million voters requiring some form of supporting document.? This should cater for the 1.6million ‘Alphanumeric’ voters and four hundred thousand ‘Document search’ voters.

The dataset below presents the same voter identification logs from the perspective of absolute numbers [CLICK IMAGE].

One can see that in absolute rather than percentage terms, Kiambu county registered the highest number of voters cleared to vote using other means. ?It recorded about two hundred thousand such voters, followed closely by Nakuru? (169K) and Meru County (154K) respectively.

The beauty of data scientists is their ability to derive meaning from huge datasets in order to inform conversations and drive policy actions.? I do hope these visualisations have contributed towards that direction.

Mr Walubengo is a lecturer at Multimedia University of Kenya, Faculty of Computing and IT. Email: [email protected], Twitter: @Jwalu

Editor's Note: The blog was updated on November 14, 2017 to state the correct form filled to clear a voter through the Document Search route is 32A, not 34A.


<var id="SFyvZPh"></var>
<cite id="SFyvZPh"></cite>
<cite id="SFyvZPh"><span id="SFyvZPh"><var id="SFyvZPh"></var></span></cite><cite id="SFyvZPh"><span id="SFyvZPh"><var id="SFyvZPh"></var></span></cite>
<ins id="SFyvZPh"><span id="SFyvZPh"><cite id="SFyvZPh"></cite></span></ins>
<var id="SFyvZPh"><strike id="SFyvZPh"><menuitem id="SFyvZPh"></menuitem></strike></var>
<var id="SFyvZPh"><span id="SFyvZPh"></span></var><ins id="SFyvZPh"><span id="SFyvZPh"><cite id="SFyvZPh"></cite></span></ins><var id="SFyvZPh"><video id="SFyvZPh"><menuitem id="SFyvZPh"></menuitem></video></var>
<cite id="SFyvZPh"></cite>
<cite id="SFyvZPh"></cite>
<cite id="SFyvZPh"></cite>
<ins id="SFyvZPh"></ins><ins id="SFyvZPh"><span id="SFyvZPh"><cite id="SFyvZPh"></cite></span></ins><ins id="SFyvZPh"><span id="SFyvZPh"><cite id="SFyvZPh"></cite></span></ins>
<var id="SFyvZPh"><span id="SFyvZPh"></span></var><cite id="SFyvZPh"><video id="SFyvZPh"></video></cite>
<ins id="SFyvZPh"></ins>
<cite id="SFyvZPh"></cite>
<cite id="SFyvZPh"></cite><cite id="SFyvZPh"><span id="SFyvZPh"></span></cite><cite id="SFyvZPh"></cite>
<cite id="SFyvZPh"></cite><cite id="SFyvZPh"><span id="SFyvZPh"></span></cite><cite id="SFyvZPh"><span id="SFyvZPh"></span></cite>
<ins id="SFyvZPh"></ins>
<var id="SFyvZPh"><video id="SFyvZPh"><thead id="SFyvZPh"></thead></video></var>
<cite id="SFyvZPh"></cite>
  • 5414891071 2018-01-21
  • 392961070 2018-01-21
  • 9593791069 2018-01-21
  • 9326521068 2018-01-21
  • 8204451067 2018-01-21
  • 3174701066 2018-01-21
  • 1649671065 2018-01-21
  • 5257251064 2018-01-21
  • 5865741063 2018-01-21
  • 6353551062 2018-01-21
  • 2642131061 2018-01-20
  • 8004871060 2018-01-20
  • 284371059 2018-01-20
  • 2901131058 2018-01-20
  • 1957611057 2018-01-20
  • 879421056 2018-01-20
  • 8217961055 2018-01-20
  • 2423161054 2018-01-19
  • 9005611053 2018-01-19
  • 8099421052 2018-01-19