# Interesting Things That You Need To Know About Statistics FYI!

Statistics is a subject which 80% of the students have known about while studying in school or college. A lot of students also up take it for higher studies and there are so many queries which they have but are often afraid to ask because they feel they might seem like an idiot in the class.

So, this post is going to solve most of your queries if not all of it. When you think about it your class grades were practically the averages of scores which you received for tests and other assignments.

Moreover, most of your classes were graded on a specific curve, which meant you required the concepts of the normal distributionstandard deviations, and other statistical terms. Now, think about your scores on specialized tests, like the SAT, which were represented in percentiles.

You must have learned about pie charts and bar graphs, scatter plots, and variousother ways to display data. I am pretty confident that you might even have learned about equations for certain elementary curves and lines. So, now that you are reading this post, you already have enough knowledge about the basic of statistics.

Statistics is omnipresent!

Well, apart from this also there are so many reasons which prove statistics is very important for each individual and is dealt with in day to day life:

• Statistics play a great role in the medical field. Every time a medical drug is launched in the market, the scientists have to show a valid rate for the effectiveness of the drug.
• You all must have an idea about insurance. Practically, everybody has some kind of insurance, whether it is medical, home or any other form of insurance. Here also, a lot of statistics is involved. Based on an individual application a lot of organisations use statistical models to calculate the risk of providing insurance.
• Ever thought about the weather forecast?Even that is done on the basis of certain statistical concepts which predict the weather beforehand.
• Scientists and researchers have been using statistical data for quite a long time now. Without statistics it would be a waste of time and money.
• Statistical data allows us to collect a lot of information which is important for day to day life. Every individual makes a lot of assumptions and predictions each of which is based on statistics.

Isn’t it amazing that without any thought process all of us include statistics in our daily life? So, next time you’re asked if you know statistics I hope your answer will be a big yes.

Focussing on the professional front

When students up take a statistics course, there are a few important terms and techniques that they have to focus on in order to complete their course with good grades. Here in this post the students will get a clear picture of the basic statistical concepts.

• Data visualization, descriptive statistics, and distributions: It is a need based analysis where the student has to understand the variance among the population.
• Group mean differences:
• Correlation and regression: They are two forms of data which depend on multivariate distribution. Correlation is more about the association between two variables whereas regression depends on predicting the value of a dependent variable on the basis of the value of the independent variable.
• Hypothesis testing: It is based on the difference between statistical significance and practical significance. It is dependent on probability as well.
• Analysis of categorical data
• Probability

Everything is uncertain

Well, to start with the fundamental difference between statistics and the other types of data analysis is that statistics is uncertain and entirely based on probabilities. The input data will always have variabilities and the results are thus always different. Every concept of statistics is based on numbers ad that is how it is predicted.

The various forms of data measurement include:

1. Natural Variability:

It is dependent on the uncertainty as well as the variability in the population patterns. It also measures the inherent difference which lies between the population and a sample. So, statistics is uncertain and can only be predicted.

1. Sampling Variability:

This is a specific set of difference which lies between the sample and the population that is dependent on the non-representative value of the sample.

1. Measurement variability:

This depends on the difference between the sample and the population and it accounts to the method how the data is measured.

1. Environmental variability:

Sometimes the data also depends on external factors and during that period the difference between the sample and the population.

The matrix is crucial

Matrices are very important when you study statistics and it is quite convenient to assemble data which is required for collation of date. After that, these matrices can be analysed by computers through mathematical calculations. Well, matrices are very much like spreadsheets. Similar to spreadsheets, matrices also have rows and columns which are individually referred to as cells.

The rows mostly represent the measured data, recording, observations, or survey respondents as well. Whereas the columns represent biological parameters, meteorological data, survey responses, and other things. The samples, variables and the data together contribute to the matrix that is essential for statistical data analysis.

Functionality of statistics

Statistics doesn’t restrict to just description and testing. It can be used for various other purposes like:

• Identification and classification
• Comparing and testing
• Predictions
• Analysis and explanations

Troubleshooting time

Well, as I mentioned before there are a lot of question which students have queries about. I have often helped them to clarify their concepts. So, here in this post I am jotting down some frequently asked questions which might come handy for the statistics students.

• How do I relate to p-value?

Often students come across p-values which are like bullets coming out from everywhere in analysis and statistical data collection. A p-value basically is required to express the probability of getting a known result from a hypothesis statistical test. It can sometimes also be a more extreme result, but only if the null hypothesis applies to it successfully.

So, if we are trying to reject the null hypothesis, then we will get to know the odds of getting our experimental data accurately if the null hypothesis is correct. Now, if the odds are sufficiently low as expected only then can we feel confident in rejecting the null hypothesis. Thus, we can look forward to accept the alternative hypothesis.

Then again a question arises that what do we mean by sufficiently low? Theoretically, the typical fixed significance level is around 0.05. So, now you must understand that if the probability showed by the p-value is less than 5% then you can reject the null hypothesis.

But, what if the fixed significance level is not accurate?

So, in case we consider 5% then why not 6% or 7%? So, remember that such data is very much arbitrary and there is no way you should throw away data if the p-value is anywhere between 6-8%!

• Which hypotheses should I consider?

The null hypothesis practically always states the status quo: which means there is no specific difference between two populations. If we consider about agriculture, then a null hypothesis will mean that there is no effect of adding fertiliser, and there will not be any relationship between weather and growth rates.

But, is that the case always? Is it acceptable?

So, generally what happens is that the scientists conduct an experiment in order to reject the null hypothesis.

They build up certain evidences, through the data collection, which would prove against the null. But, only if the evidence is sufficient only then, one can say with a certain degree of probability that the null hypothesis can be rejected.

Once, the null hypothesis is rejected then we can finally accept the alternative hypothesis. This hypothesis will obviously state the opposite of the null that there is a difference, there is an effect, so there is an appropriate relationship.

• What is the best technique to analyse data?

This is one of the most common queries that I have got from the students so far. It is very complicated for them to figure out which method is appropriate for their analysis. So, in order to start the analysis it is crucial for you to understand the type of variable that you’re dealing with. These variables belong to two types namely: continuous and categorical variables.

Now, continuous variables can have any value. For example, say you are to measure a certain time that is required for completing a reaction, so the result could be 30 seconds, one minute, or even two minutes. So it’s simply a random value which can be recorded for use.

Categorical variables are the most easy to predict ones. For example, there might be four different field areas, or three different brands of fertiliser.

So, this would be easy to predict and analyse. Also, you should remember that all continuous variables can be converted into categorical variables but not vice-versa! So, it is easy to record data as continuous for more options as available and then convert it to categorical if needed.

• Which analysis tools should be used for better results?

Once you are clear about the technique, then you have to focus on the tools. The two most convenient and easy to use statistical analysis tools are ANOVA (analysis of variance) and linear regression.

So, ANOVA is quite an interesting analysis tool which is used to compare continuous and categorical variables like the use of quality fertilisers as compared to the increase in the growth of the plant in centimetres. Similarly the linear regression tool is used for comparing two continuous variables like growth vs. time!

So, once you know the variables that you are working with and are determined to use the tool then you are all set to analyse that huge data coming your way. Trust me it won’t take more than a few hours to analysis one full spread sheet.

Fast-forward to some hacks

Well, the more you practice the sums, the better you’ll get on data analysis and solving statistical problems. Here are some of the most assorted handy tips which I have been personally using and recommending to my students.

1. Work as many problems as you can, and do no waste much time on one problem.
2. Start studying weeks before the exam if you want to score well.
3. Use several types of learning which will include terminologies as well as formulas.  Memorizingdefinitions is extremely important, but you must not stop there. It’s time to go further and explore all the options.  You must ensure whether you can analyse a certain problem and understand about you strengths and weaknesses.
4. Gather information from various resources especially from the internet as well as through e-books. It is always a bonus to have pre-prepared notes at least a week before the exam because that will not only help you to remember things clearly but will also shorten your memorizing time.
5. Last but not the least you must take help from the tutors and mentors because they will surely clarify your concept.

Now, I know these are some of the simplest tips which you might have heard ever but trust me incorporate these hacks in your regular schedule and you are bound to see the difference in your result. Statistics is an extremely interesting subject and remember that it is a part of our daily lives. Without statistics life won’t really make any sense. So, keep exploring and analysing.

Author bio:

Michelle Johnson is one of the most reputed statistics professors and has been guiding students for more than six years now. She is very friendly and has unique hacks up her sleeves. Her students always look up to her for expert tips in economics. She is also quite a popular blogger and has been working with a lot of renowned companies as well. Her expertise lies in correlation, regression, probability, financial management, and ANOVA techniques.