According to Mendenhall and Beaver (2011), descriptive statistics is the analysis of data that helps to summarize, show and describe data in a way that is more meaningful. In this regards, it provides a concise summary of data. This data can be graphically or numerically summarized. Using graph or pie charts in descriptive statistics helps one interpret data collected in a much easier way than providing the same data in a wide amount of raw data. Descriptive statistics use mean, median and mode as measures of central tendency and use standard deviation and variance as measures of dispersion.
Inferential statistics, on the other hand, uses data collected randomly in a population (samples) and later on making inferences about the population in question. Inferential statistics are quite valuable in a situation where it would be impossible to examine each member in a population, for example, acquiring data from the whole world about their preferences between tea and coffee. Inferential statistics, therefore, allows us to use these samples to make the generalization about the population where the samples were drawn during data collection. There are two forms of inferential statistics:
The application of estimation statistics is based on the determination of values of the population where the data is being collected by the researcher. For example, when collecting data about how many people would prefer tea over coffee in a population representing the whole world, in a case where 35% of the population would prefer the tea, using inferential statistics, it would be safe to indicate that the 35% would represent the world population.
Hypothesis Testing
This technique assists researchers in drawing conclusions in a study. However, in this regard, the researcher uses tests such as t-testing to determine whether the hypothesis is true or not. The main difference between the two types of science studies, descriptive, and inferential statistics, is that the former quantitatively describes features of a data sample while the latter makes inferences about a population from which samples were drawn.
While researching data from a population by using either descriptive or inferential statistic, a quantitative form of study is more important than the qualitative part of the data since the two forms of statistical data collections tend to deal with numbers of the data in a population so as to develop quantitative feedback rather than that based on quality of the data.
The p-value is the probability of results of a statistical test where studies of a hypothesis have not yielded any positive or negative results (Taylor, 2012). These hypotheses can either be rejected or retained by the help of the statistical tests. The probability will measure the evidence against the hypothesis. If the p-value is small, that will correspond with strong evidence and vice versa.
Data obtained while obtaining the value of p will yield one of two results, either the hypothesis is rejected, or the hypothesis cannot be rejected. P-values give the difference from a previously specified statistical level. Additionally, they are also very clear and can be judged whether a value is greater or less than a previously specified limit, and this helps in determination of a rapid decision in the assessment of whether a value is significant or not.
However, this type of statistical testing has had its fair share of heavy criticism due to its shortcomings and misinterpretations. In this regard, the p-value measure is often misinterpreted and misunderstood as it fails to provide adequate data and information to the reader (Taylor, 2012). For instance, it does not explain the amount of observed effect and also the importance of the research’s outcome.
References
Mendenhall, W. & Beaver, R. (2011). Introduction to probability and statistics. Belmont: Brooks/Cole, Cengage Learning.
Taylor, G. R. (2012). Integrating quantitative and qualitative methods in research. Lanham: University Press of America.
