Data Representation Models
We live in an information age, where the volume of data processed by organizations increases exponentially. Managing this plethora of data is very crucial as even small discrepancies in data can make a significant difference to a company’s bottom line. Quality of the data reflects in the quality of decisions and hence make and break the future of any data driven system. Keeping this in mind a management role demands strong skills in understanding and using data. DI section checks our comfort level with data and data handling. Lot of data will be presented as bar charts, pie charts, tables… Understanding when & why behind these models make our life much easier while tackling DI.
There are two types of data
Quantitative: Data that can be measured. It deals with Quantities.
E.g. Alok’s height is 1.9 m
Qualitative: Data that can be described but not measured. It deals with Qualities.
E.g. Alok is tall.
Various representation models are used to present data.
A table is a means of arranging data in rows and columns. Table helps us in presenting quantitative data in a neat and extendable format.
Given below is the table that contains the score details of test takers from a test.
Here, a lot of data is given, but there is no confusion regarding which means what. We can easily traverse through rows and columns to get the required information. Say if we need to see Verbal score of David or DI score of Arun we know where exactly to look into. Also if we need to add 5 more students or 2 more sections, we can do so without making the table clumsy.
Usually we use Table when we don’t want to strip the available data. We will present complete data from which lot of information can be deducted. From this table it is difficult to get the trend or max/min values. But it is the best method when we deal with quantitative data.
A common application of tables is in our cricket score board. Using the below score table we can get the performance of EACH batsman and bowler but we may not be able to get the trend in scoring, like when the team scored most runs, when team lost most of their wickets etc…
Sometimes (or most of the times) we may not need the complete set of data. We will be interested in “Snapshots” of particular areas or trends. In such cases we will use qualitative methods of data representations which are visual in nature. You can just SEE than traversing. We will now explain various qualitative data representation models.
Bar graphs are usually used to display data that fits into categories. A bar chart is made up of columns plotted on a graph. The columns are positioned over a label that represents a categorical variable. The height of the column indicates the size of the group defined by the column label.
A “Snapshot” of our test score table is given below. We can easily SEE who scored max in quant, who scored least in verbal, who has a good distributed score in all sections and so on.
Problem with bar chart is if you add more test takers or include additional sections the chart becomes clumsy.
Coming back to cricket, a bar chart is used to show runs scored in each over for the match.
A pie chart is a circular chart divided into sectors, which represent the magnitude of subsections relative to the whole. The arc length, central angle and area of each sector are proportional to the quantity it represents.
A Pie chart depicting the sectional scores relative to the total score of Arun is given below. Here also we can SEE where Arun scored most and where he scored least.
Limitation of Pie chart is that if we need the score distribution of another person we need another Pie chart. Pie charts are widely replaced by bar charts in meeting rooms now.
This pie chart shows the run distribution in a bowling spell. So Brett Lee is a good bowler? :)
What if we want to see the trend of scoring of students in a number of tests?
A line graph is most useful in displaying data or information that changes continuously over a reference entity (like time).It helps to identify the trend by visual inspection. Given below is a line graph that plots the overall score of each test takers for five mock tests. It is not that easy to get the actual scores but we can easily SEE the scoring trend of each candidate.
In cricket we see line graphs to be used to display the scoring trend of each team
From this graph we may not be able to get a detailed picture regarding the match as we got from a score table. But we can easily see the trend of scoring of each team.
There are various other representing methods like histograms, scatter chart, Radar chart etc… but considering the types we expect in our exams it is ok not to go deep into them.
Now the most important part is how knowing these charts will help us to score better in DI.
We are used to solving questions in order but the best way to score in aptitude exam is to run through the section pick the easy questions first, solve them as fast as possible, pick the moderate ones, finish them and then come to tougher ones and solve as much as you can. This is kind of straight forward in quant and verbal where mostly we can gauge the complexity of a given question quickly. But this gets tricky in DI. We cannot say a given set is tough or easy based on the amount of data presented before us. Complexity of a DI set is directly related to the effort we need to get what is asked from what is given. There are no hard and fast rules in determining the complexity of a given DI but some common observations are as below
If a DI set presents data using quantitative tools (table) and ask you to do quantitative analysis (basically to play with the numbers), see how much effort is required for each calculations. Approximation plays a crucial role here. Sometimes there may be a question or two in the set which are unreasonably demanding in terms of required effort. Tackle the easy ones in the set and park the time consuming ones for later (if time permits).
If a DI set presents data using quantitative tools and ask for qualitative analysis (like identifying trends), It mostly come with lot of calculations at various levels. We may have to derive some values from the given data and then operate on those values again to reach the solution. It is important that we do this one carefully because if we make a mistake at some stage, our final answer will be skewed and we will never know. I suggest park these questions for the end unless you see some way to tackle them faster.
If a DI set presents data using qualitative tools (Bar graph, Pie chart …) and ask for quantitative analysis, it may get tricky picking the values from the plot and chances are more for mistakes if we don’t do it carefully. Effort required for this type of question depends on our comfort level in plotting the data from the graphs and also on the amount of data required to solve the questions.
If a DI set presents data using qualitative tools and ask for qualitative analysis, usually we can get the answer by visual inspection alone.
Happy Learning :)