Data Quality is not a Concept Now
I used to say during my research days, if you put garbage in to a computer garbage comes out from the other end.
This invariably happens.
I used to detest meta analysts.
They are the ones who takes sets of data from others and create a new set of data and assume that the new set is their own creation.
This assumption is a dangerous preoccupation.
It is like religious beliefs.
In reality they have not contributed a single bit of data. So, I collected each piece of data myself and double checked every piece.
No secretary was involved.
I was the secretary.
I was the data collector.
I was the data analyst.
It takes time but reliability stands out.
In big business these three are done by three different guys or girls and a single error is multiplied by three.
Multiple errors leads to multiplex and no validity. This is where I am against the AI or artificial intelligence. If you feed inaccurate and inadequate data the outcome is complete nonsense.
In Reality, I was against the medical data formats. An accurately kept Bed Head Ticket with a ongoing problems and outgoing problems summarised, ideally on weekly basis and if not monthly basis is adequate.
A database is only good enough for retrieving and sorting out two or three cases with identical names.
Nothing more.
The rest becomes redundant statistics.
This does not underscore a simple database kept by a GP with regular contacts of subjects. Unfortunately, most GPS become data collectors and referral agents of substandard quality.
I was obsessed about a database at birth in my research work but my attempt at extending beyond birth was futile since the responsibility of tying the knot was delegated to midwives (mind you, there were a few good ones) who were poorly trained about accuracy, as regards to the biodata collected and recorded on BHTs and the nurses are busy with birth certificates and rest of the clinical work. The final birth record book was simple but had many errors including the sex of the baby.
Mind you, I cracked the nut.
The global belief is that male babies are bigger and healthy but in reality the womb of the mother is non discriminatory to the sex of the baby, if Body Mass Indexes (BMII) are compared with the sex (male babies are bit heavier due to slight increase in longitudinal scale) of the baby there is no difference irrespective of the sex .
There was no sex difference except the slight excess of death of male babies of small mothers, who worked until the onset of parturition. Especially so of mothers who were essentially labourers. The remedy is simple no work after 35 weeks of pregnancy and plenty of sleep and good food.
Iron treatment had no relationship and only two mothers were anaemic (excluded from the main database). But B vitamins especially soluble ones are important and until 6 months after stopping breast feeding. The lactating mother is often neglected and baby is the centre of attraction globally.
I used to say during my research days, if you put garbage in to a computer garbage comes out from the other end.
This invariably happens.
I used to detest meta analysts.
They are the ones who takes sets of data from others and create a new set of data and assume that the new set is their own creation.
This assumption is a dangerous preoccupation.
It is like religious beliefs.
In reality they have not contributed a single bit of data. So, I collected each piece of data myself and double checked every piece.
No secretary was involved.
I was the secretary.
I was the data collector.
I was the data analyst.
It takes time but reliability stands out.
In big business these three are done by three different guys or girls and a single error is multiplied by three.
Multiple errors leads to multiplex and no validity. This is where I am against the AI or artificial intelligence. If you feed inaccurate and inadequate data the outcome is complete nonsense.
In Reality, I was against the medical data formats. An accurately kept Bed Head Ticket with a ongoing problems and outgoing problems summarised, ideally on weekly basis and if not monthly basis is adequate.
A database is only good enough for retrieving and sorting out two or three cases with identical names.
Nothing more.
The rest becomes redundant statistics.
This does not underscore a simple database kept by a GP with regular contacts of subjects. Unfortunately, most GPS become data collectors and referral agents of substandard quality.
I was obsessed about a database at birth in my research work but my attempt at extending beyond birth was futile since the responsibility of tying the knot was delegated to midwives (mind you, there were a few good ones) who were poorly trained about accuracy, as regards to the biodata collected and recorded on BHTs and the nurses are busy with birth certificates and rest of the clinical work. The final birth record book was simple but had many errors including the sex of the baby.
Mind you, I cracked the nut.
The global belief is that male babies are bigger and healthy but in reality the womb of the mother is non discriminatory to the sex of the baby, if Body Mass Indexes (BMII) are compared with the sex (male babies are bit heavier due to slight increase in longitudinal scale) of the baby there is no difference irrespective of the sex .
There was no sex difference except the slight excess of death of male babies of small mothers, who worked until the onset of parturition. Especially so of mothers who were essentially labourers. The remedy is simple no work after 35 weeks of pregnancy and plenty of sleep and good food.
Iron treatment had no relationship and only two mothers were anaemic (excluded from the main database). But B vitamins especially soluble ones are important and until 6 months after stopping breast feeding. The lactating mother is often neglected and baby is the centre of attraction globally.
No comments:
Post a Comment