Thumbprint verification

The ongoing hullabaloo over voter thumb verification using magnetic ink during the last general elections has made it difficult to separate fact from fiction. Every Tom, Dick and Harry claims to be an expert on thumb forensics. Biometric experts are sitting on the sidelines and waiting to be asked about their observation on scientific knowledge and rational logic.
No finger printing device or ink is foolproof or can give 100 percent accurate results. Handprint-based security systems all over the world have to constantly update the database of staff so that fingerprint records are always in sync. Human fingerprints change with age and due to the nature of work. A typical example was the recent PTA-mandated SIM verification process. The majority of senior citizens – including my father who is 78 years old – had to get their SIMs transferred to other family members because Nadra machines could not verify their fingerprints. Fingerprints of Nicop-holding Pakistanis are outdated in the Nadra database. They are being asked to get new fingerprints either from embassies or overseas Nadra offices. Those respectable judges in charge of the judicial commission and election tribunals may be legal experts but not forensic or biometric experts. They are well-advised to make decisions based on legality and not on fictitious political propaganda.
Asad Qureshi
Islamabad
*****
This refers to the news report, ‘Nadra system fails to verify 93,852 thumb impressions’ (May 10). According to the report, out of 184,151 votes, 73,478 (40 percent) could be verified. What about the remaining 60 percent votes? The report correctly describes this as a failure of the Nadra system rather than large-scale bogus voting. However, rival political parties are interpreting the results differently to their advantage and confounding the confusion.
My personal experience indicates that Nadra’s system of verification of thumb impressions is not foolproof. For biometric

verification of my SIM card, my thumb impression was not recognised at one company’s office. I tried at another mobile company where it was accepted. There is a need for what statisticians call ‘controlled experiments’ by experts in the field to determine the level of confidence that can be put on Nadra’s results. Similarly, 6,123 invalid CNICs (3.3 percent of the total) indicated in the report could be errors of transcription rather than bogus voting. In the push and pull of voters around election staff, recording of 13 digits of CNIC number is liable to transcription errors.
Abdul Majeed
Islamabad