An overview of how systems store and retrieve data
As a system and data practitioner, I used to be puzzled why people would, when asking for reports, describe easy requests as hard and hard requests as easy.
1. “Hi Adam, another report, please, and I’m sorry that it’s a bit of a monster. Could I have a contact list of all current members of staff?”
2. “Hi Adam, I’ve got a quick one for you. Could you get me a contact list of all women currently employed in the central office who are between 45 and 60 and earn less than 45K?”
For someone well versed in data queries, the first query is infinitely easier than the second, but so many colleagues seem to regard the second as the easy one. I was always confused as to why this was until eventually, the scales fell from my eyes and I realized that we were looking at data storage in two different ways.
So what’s the difference?
My realization was that most of my colleagues view HR records through a traditional, paper-based eyes; they think of the data in the system being organised as if they were paper files. My apologies to any reader who already knows, but that means that in an organisation of 5,000 people there will be a cupboard or cupboards with 5,000 folders, one per employee, organised A-Z. Within each folder will be one or more pieces of paper. These will be related to aspects of the person’s period of employment, but for the purposes of this explanation I’ll restrict myself to three pieces of information that are ubiquitous:
· Personal Data: this is data that will almost never change over the course of employment:
o Date of Birth
o Address
o Government ID number (Social Security, National Insurance, etc)
o Contact details
· Contractual: specific data related to the job. This can change over time due to transfers or promotions:
o Job Title
o Work location/office
o Start Date (and end date if Fixed)
o Remuneration
· Absences
o Sickness
o Holiday
which would give us 5,000 folders with 3 (assuming everyone has had a least one instance of absence) or more pieces of paper, a minimum of 15,000 individual pieces of information.
So how does a system differ?
In order to understand how a computer stores info, I’ll ask the reader to engage in a mental exercise. To start with, please imagine that all 5,000 paper-based files are accessible
· From each of the folders, remove all the personal information pages and put them (in no discernable order) into one folder
· Then, remove all contractual information and put into another folder as in the previous step
· Finally, move all the absence information and put into a third, separate folder
So instead of 5,000 folders with at least 3 pieces of paper in each, we now have three, large, folders with a minimum of 5,000 pieces of paper in each.
It’s important to note that these three folders also keep the details of all former employees, unless specific steps are taken.
How does this affect reporting?
In a paper based environment to get the contact information for the first report, we would have to look up each person and copy their contact info… 5,000 times. For the second, we would need to do the same, but for only a handful of records.
For the second (albeit electronic) system, the report would be generated:
· Search all the contractual files for current members of staff
o This would, incidentally, take care of the first query: one-step=easier
· Search current records for females
· Search current, female records for central office
· Search current, female, central records for under £45K
In the end, the second would be faster than looking up each record individually, due simply to the incredible speed that the computer can search, but the parameters are more involved.
Comments