Types of Statistics

 There are two basic types of statistics; descriptive statistics and inferential statistics.

(i)     Descriptive statistics: This consists of methods for organizing, displaying, and describing data by using tables, graphs, and numerical measures. Most of the statistical information in newspapers, magazines, company reports, and other publications consists of data that are summarised and presented in a form that is easy for the reader to understand (i.e. Descriptive form). Descriptive statistics are used to summarize results of research or to help the teacher give the best description of examination results in the school. For example, the Grade 12 results for 2011 were the best at David Livingstone High because for the first time the school recorded 25% of those who got 10 points and below.

(ii)   Inferential Statistics: consist of methods for drawing and measuring the reliability of conclusions about population based on the information obtained from a sample of the population. For example, we may make some decisions about the political views of all college and university students based on the political views of 1000 students selected from a few colleges and universities. Inferential Statistics are used to arrive at general conclusions from research results and to test hypotheses. Inferential statistics could be used by the teacher to draw general conclusions from test results in a particular subject about how the performance of learners could be in that subject during the final examination.

Importance of statistics

Statistics tell us how often something happens and in education, statistics are used among other things:

(a)  To help educators summarise educational results using the most desirable options during the teaching and learning process.

(b)  To manipulate educational results, educators need to be knowledgeable about the vocabulary, concepts, and statistical procedures used in statistical studies.

(c)  Statistics are also used to help describe the results of studies and to reach feasible conclusion based on the results.

Limitations of Statistics

Statistics are not a pane seer to all problems. Statistical methods have a limitation when certain problems cannot be quantified, such as happiness, sadness, love, depression and so on. However, there are some qualitative methods which can be used to analyse situations which can’t be quantified. Below we discuss a few limitations:

·         Qualitative aspect ignored: The statistical methods don't study the nature of phenomenon which cannot be expressed in quantitative terms. Such phenomena cannot be a part of the study of statistics. These include health, riches, intelligence, happiness, sadness, love, depression etc. It needs conversion of qualitative data into quantitative data.

·         Results are true only on average: Usually Statistics deals with only aggregates of facts or items and it does not recognize any individual item. Further, the results are interpolated for which time series or regression or probability can be used. These are not absolutely true.

·         It does not depict the entire story of phenomenon: This is when phenomenon happens that is due to many causes, but all these causes cannot be expressed in terms of data. So we cannot reach at the correct conclusions. So we analyse only the data we find quantitatively and not qualitatively.

·         It is liable to be miscued: One of the shortcomings of statistics is that they do not bear on their face the label of their quality. Data may have been collected by inexperienced persons or they may have been dishonest or biased. So data must be used with a caution. Otherwise results may prove to be disastrous.

·         Laws are not exact: As far as two fundamental laws are concerned with statistics, law of inertia of large numbers and law of statistical regularity are not as good as their science laws as  they are based on probability. So the results will not always be as good as of scientific laws. Here only approximations are made.

·         Too many methods to study problems. In tins subject we use so many methods to find a single result. Variation can be found by quartile deviation, mean deviation or standard deviations and results vary in each case.

·         Statistical results are not always beyond doubt: "Statistics deals only with measurable aspects of things and therefore, can seldom give the complete solution to problem. In short, they provide a basis for judgement but not the whole judgment.

Role (Function) of a Statistician in Research

·         Acquisition (collection of data): The primary concern of a statistician is acquisition of data using either sample surveys or experiments. In doing this, the statistician has to make a decision on the survey procedure if they are using a sample survey. Hence the statistician needs to determine; the type of data need to be collected, the data collection techniques, the sample size need, sampling methods and so on.

·         Selection of the best method for making inferences: Once data has been collected and run through the computer, the next thing is to choose the method for making inferences. Depending on the sample size and whether a probability sampling was used or not, difference inference methods would be chosen. There are two types of tests which are usually employed.

                (a)      Parametric Tests: These tests assume that you want to say something about the population parameter based on the sample drawn using probability sampling methods. These tests are exclusive to quantitative type of data.

               (b)      Non-parametric Tests: Nonparametric tests are sometimes called distribution-free tests because they are based on fewer assumptions (i.e. they do not assume any probability distribution of an outcome)

Determination (Evaluation) of the goodness of the inference: the interest is to try and establish how accurately a sample value estimates the population value. If you are using probability sampling, the sample statistic will not equal the true population parameter, hence there is need to determine the margin of error. That is how close it is to the real value.