#what is data analysis

Data analytics helps forecast future value whenever we want to know what will happen next. To draw conclusions from data, one uses data analytics. It often entails data collecting and examination and has one or more users. 

Data analytics helps forecast future value whenever we want to know what will happen next. To draw conclusions from data, one uses data analytics. It often entails data collecting and examination and has one or more users.  Data analysis is performed on the historical dataset, to comprehend what has occurred thus far from the data. In short, data analysis is used to extract precise insights from…

                                       Image source:  aircloak.com

In the current scenario, there has been an exponential increase in the amount of data available. Considering the manifold increase in both unstructured and structured being generated from innumerable resources viz. humans, documents, organisations, different electronic media (mobile , laptops/desktops), software applications, print media, it is of vital importance to extract the desired information from it.

One needs to draw useful and informative insights from the raw data collected which could either be homogeneous (systematic) and heterogeneous (unsystematic). In addition, the data collected can be of two types:

Qualitative (defining the attributes/ characteristics, not numerical)

Quantitative (Numerical)

Once the data is collected either from a primary or a secondary resource, it should be organised, tabulated and made more comprehensive. Presenting the data in a meaningful manner helps to analyse it and draw inferences.

In statistics, data analysis is bifurcated into i) Descriptive Statistics ii) Exploratory Data Analysis (EDA) and iii) Confirmatory Data Analysis (CDA). Data analysis is a multistage process.

The next step after data collection is data cleaning i.e. to understand the data and figure out if a pattern in the data could be observed so that the analysis becomes easier. In context to quantitative methods, the data set should be cleaned to remove outliers (something exhibiting a completely different characteristics) which could cause data anomaly. Further, with the help of different spell checker and grammar correcting software one could also reduce the number of mistyped words and grammatical errors. Reasons for figuring out errors in the first step are to avoid faulty analysis at the later stages.

Quality analysis involves calculating the frequency counts, descriptive statistics (Mean, mode, Median, Standard Deviation), skewness (the alignment of the data, direction right or left) , kurtosis (Peakedness of the data i.e. either  platykurtic, mesokurtic and leptokurtic).

Quantitative Data Analysis comprises of the following methods:

Descriptive Analysis: Descriptive analysis is the first level of analysis which helps the researchers to understand the variables and summarize them. It includes finding out the mean, median , mode, frequency, range.

Inferential Analysis: In Inferential analysis the researcher aims to generalise results and make predictions on the basis of correlation (relationship between two variables), regression (cause and effect relationship between two variables), Analysis of Variance (ANOVA, helps to figure out the extent to which two or more groups differ).

Understanding Qualitative data

One needs to follow the below mentioned steps to perform qualitative analysis:

Familiarity with the data: Read the data multiple number of times to understand what it is all about. This also involves transcribing the data.

Understand Research Objectives: Every research problem which is formulated has an objective, henceforth it is of prime importance to very well understand the research objectives.

Develop a research framework: Discover the broader ideas, concepts, behavioural patterns and allocate them codes. This helps to make the data more structured.

Identify for similar patterns: Once all the aforementioned steps are followed, the researcher should find out connections and links in the given data. Patterns in qualitative data could be figured out either by 1) Word-based methods (Word repetitions) or 2) Scrutiny based techniques

Word based techniques simply means that the researcher re-reads the text and identifies the words used more often. For instance a researcher is conducting a study on Indian youth and politics, wherein the researcher discovers that the most commonly used words are “greed”, “corruption” and uses them for analysis.

Compare and contrast method is one example for scrutiny based technique. The researcher tries to comprehend the thematic representation of data. For example, one theme could be: The importance of a counsellor in school. Here the data could be divided into those who think there should be a counsellor and those who don’t.

Thus qualitative data helps the researcher to look for analogies, connections and metaphors in the data.