More reason not to worry but say hello to stats!

  To be able to use any body of knowledge we need to develop a unique outlook towards it. It is important to develop a statistical thinking so to say to be able to successfully understand and apply statistics in your work/research. To cut a long story short, you need to have a starting point of view. 

     Statistics is the science of data. Period. It is about collecting, classifying, summarizing, organizing, analyzing and interpreting numerical information. It is applicable in most of the disciplines both scientific and non-scientific. So if we start thinking in data terms, it helps a lot in understanding and applying statistics. Statistics is nothing but a set of concepts and rules to apply those concepts for deriving meaning out of data. So the starting point is data. Every statistical problem or scenario ultimately boils down to data. We all have learnt about data at school. However, as mentioned in the previous post, the understanding of data is somehow lost in the conceptual knowledge of  different subjects that are being emphasized in school. 

     Like any other discipline statistics helps solve problems through measurement. Though measurement does not answer all questions it does help us gain better understanding of most situations, which gives us more control over problems, a better handle to overcome them. 

For instance, 

1) You as a medical researcher might want to determine the efficacy of a particular treatment or drug on a given ailment or medical condition. Herein you might want to measure the effect of a given treatment on a specified number of patients suffering from a given condition.  This can give us a fair idea of the effectiveness of the given treatment. For this we need to collect data about the patient conditions, the dosage and mode of the drug/treatment, its timing and effects on each patient.

2) You as a researcher in Human Resources might want to measure the effect of a specific motivational plan on the employees in particular sector. You will collect the employee data as well as the data regarding the effect of the motivational plan on them.

3) As a market researcher you might want to establish the effect of an advertising program on the sales of a given brand or company. Herein you need the sales data prior to and post the running of the advertisement.

4) As a financial analyst you might want to collect financial data to measure the effect of a given investment action like buy/sell/hold on a specific stock or a portfolio of stocks over a period of time. You need the stock returns data for the said time period.

5) As a scientist on NASA's lunar program you might want to analyse the fuel consumption in different stages of a lunar mission to optimize the fuel consumption on the next mission.
You need data regarding the consumption of fuel in the different stages of the previous lunar missions.

     The examples can be numerous and you can be overwhelmed by the sheer variety of situations where statistics can be applied, but an interesting starting point can be a set of these simple questions-

1. What is the purpose of the research?
2. What is the data to be collected? 
3. Where to find it ?  
4. How to collect it?

     At this point it is important to understand that data does not always necessarily has to be measurable. There is something called qualitative data as well, which leads to information that is not measurable but is important and crucial nevertheless.

Watch video at following link to understand what I mean..

Data collection overview

     A good research has to consider all the data collection methods that might be relevant and crucial to the purpose and constraints of your research.

Since we are particularly concerned about statistics here, it is important to acknowledge the role of qualitative data in overall research and how it can useful to the statistical analysis that might required in your research.

Qualitative data often complements and supports the quantitative data measurement and statistical analysis of the same.

For instance in above examples,

1) In the medical research example, you may want to collect qualitative data about the signs, symptoms, experiences and related medical conditions of the patients.

2) In the market research example, you may want to explore the brand preferences of the customers to arrive at a starting point for defining the variables for quantitative research.

3) In the Financial analysis example, the scope of qualitative data is very limited, however nowadays a relatively new branch of finance namely behavioral finance might require the analyst to understand the human instincts of greed, fear and hope in a given context of the financial markets where the investment actions are being studied.

4) In the NASA example, the experiences of the astronauts can yield important information about optimization of fuel consumption in different stages as survival in space can be an important balancing or trade off factor for fuel consumption.

In the next post, we will talk more about data collection, till then try and ponder a little about what has been discussed so far...

Math and Statistics relationship - relevance to research aspirants

Statistics and Math: the link and why should you not worry about it..

Statistics is taught in schools as part of the math syllabus and thus it is usually perceived as a mathematical concept. The schools emphasize on computations of complex statistical information rather than development of sound analytical skills and points of view (Gattuso, 2006). This usually makes the students perceive statistics as a dry and number crunching discipline. Nothing could be farther from truth. Though statistics does involve numbers crunching in the form of data processing as it revolves around data and how it can be made to make sense for common people, it does so in a very smooth, sophisticated yet logical manner. So if you are good at logic, you can be really good in statistics. The typical calculations are anyway nowadays done mostly by various statistical software which aid in arriving at patterns of data that can be interpreted in a useful manner.

     All said and done, all statistical concepts do have some or the other underlying mathematical principles to them. However, as we begin to appreciate the applicability of the same to our day to day lives, it all starts falling into place. Things start becoming clearer and clearer and if we do not lose track or get distracted we can soon reach a point where we start seeing data and numbers as a source of informed opinion on things we earlier thought we could never have an opinion.  Unfortunately, the misconceptions about statistics picked up at school level are carried forward to higher levels of education where research and often quantitative research becomes an important means to prove your proficiency in the given subject. While research does not always involve statistics, as there are many different kinds of qualitative approaches to research, the level of logical or reasoning ability required in both qualitative and quantitative approaches is unique in its own way.

     So if you are pursuing academic research in any form, be it a school project, a research paper, a coursework, a case study, a dissertation or a PhD thesis, my advice is not to avoid quantitative research in the form of statistical thought and analysis as this might be the preferred and appropriate way to go about in many research situations. Usually qualitative and quantitative research complement each other and we can ignore one only at the cost of other and indeed consequently at the cost of good quality research.

     Statistics can in fact be treated as mathematics applied in an interesting and motivating manner. Numbers are found in everyday life and most of the mathematics included in statistics has already been taught at school level. Rest is only well reasoned concepts and formulas based on the math that has already been studied in school.

    There are so many interesting uses of statistics we can see all around us. The election opinion polls results we see during election times are based on statistics, the weather reports on TV, the airlines flight timings, the expected arrival and departure times of trains, the efficacy or composition of the medicine you might have last consumed are all based on statistics. The performance of your favorite sports star is analysed using statistics and presented to you. The economic performance of your country including inflation, GDP, per-capita income all are calculated and interpreted using statistical principles.

So, the next time you frown at the prospect of using statistics in research, think twice. We will discuss more in forthcoming posts of this series about how to develop a simplistic yet effective understanding of statistical concepts, so that you can pursue your research confidently.

References

Gattuso, L., 2006. Statistics and Mathematics: Is it Possible to Create Fruitful Links? [Online] Available at: https://pdfs.semanticscholar.org/78c4/f1509602750b2f1c1d3a6db8c66b38109af0.pdf [Accessed 2017 July 2017].