1
INTRODUCTION TO DATA TYPES AND SPSS OPERATIONS

LEARNING OBJECTIVES

After completing this chapter, you should be able to do the following:

• Understand different data types generated in research
• Learn the nature of variables
• Know various data cleaning methods
• Learn to install SPSS package in computer
• Prepare data file in SPSS

1.1 INTRODUCTION

Due to large stake involved in sports, research in this area is gaining momentum in different universities of the world. Even developing countries have started introducing sports sciences in different universities. The sole purpose is to create specific knowledge required for enhancing sports performance. Everyday, enormous data is being generated in the area of sports all over the world, which can be used to draw meaningful conclusions. Scientists have started organizing experiments by taking athletes as subjects. It is therefore required to support these scientists with analytical skill set to carry out their business. Since they deal with the data, it is essential that they are aware of its nature. Depending upon the data types, one identifies the relevant analytical technique for addressing research issues. Sports research can broadly be classified into two categories: descriptive and analytical. In descriptive research, the nature of dataset is investigated from different perspectives. Several statistics like mean, standard deviation, coefficient of variation, skewness, kurtosis, and percentiles are used to describe the characteristics of the dataset. Many interesting facts about the population can be investigated by using these descriptive statistics. Analytical research broadly follows two approach: exploratory and confirmatory. In explorative research, focus is on discovering the hidden relationships. It is done by hypothesis testing, data modeling, and using multivariate analysis. On the other hand, in confirmatory studies, some of the facts are either confirmed or denied on the basis of hypothesis testing.

Numerous statistical techniques are available to the researchers for analyzing their research data. Selection of an appropriate technique depends upon the research questions being investigated in the study. Due to complexities of different analytical solutions in sports research, one needs to use some user-friendly software package. This chapter will acquaint you with different types of data that are generated in sports research and some of the widely used statistical techniques by the research scholars to solve them for answering different research questions by using the most popular IBM SPSS Statistics package.

1.2 TYPES OF DATA

It is essential to know the types of data generated in research studies because choosing statistical test for analyzing data depends upon its type. Data can be classified into two categories: metric and nonmetric. Metric data is analyzed by using parametric tests such as t, F, Z, correlation coefficient, etc., whereas nonparametric tests such as Wilcoxon signed-ranked, Chi-square, Mann-Whitney U, and Kruskal Wallis are used to analyze nonmetric data.

Parametric tests are more reliable than the nonparametric, but to use such tests certain assumptions must be satisfied. On the other hand, nonparametric tests are more flexible, easy to use, and not many assumptions are required to use them.

Nonmetric and metric data are also known as qualitative and quantitative data, respectively. Nonmetric data is further classified into nominal and ordinal. On the other hand, metric data is classified into interval and ratio. These classification is based on the level of measurements. The details of these four types of data have been discussed under two categories: qualitative data and quantitative data.

1.2.1 Qualitative Data

Qualitative data is a categorical measurement and is expressed not in terms of numbers, rather by means of a natural language description. It is often known as "categorical" data. For instance, smoking habit?=?"smoker" and gender?=?"male" are the examples of categorical data. These data can be measured on two different scales: nominal and ordinal.

1.2.1.1 Nominal Scale

Variables measured on this scale are known as categorical variables. Categorical variables result from a selection of categories. Examples might be response (agree, disagree), sports specialization, race, religion, etc. If in a class 30 subjects are male and 20 are female, no gradation is possible. In other words, 30 do not indicate that the males are better than the female in some sense.

1.2.1.2 Ordinal Scale

Variables that are assessed on the ordinal scale are also known as categorical variables, but here the categories are ordered. Such variables are also called "ordinal variables." Categorical variables that assess performance (good, average, poor, etc.) are ordinal variables. Similarly, the variables that measure attitude (strongly agree, agree, undecided, disagree, and strongly disagree) are also ordinal variables. On the basis of the order of these variables, we may not know the magnitude of the measured phenomenon of an individual, but we can always grade them. For instance, if A's playing ability in soccer is good and B's is average, we can always conclude that the A is better than B, but how much is not known. Moreover, the distance between the ordered categories is also not same and measurable.

1.2.2 Quantitative Data

Quantitative data is a numerical measurement expressed in terms of numbers. It is not necessary that all numbers are continuous and measurable. For instance, the roll number is a number, but not something that one can add or subtract. Quantitative data are always associated with a scale measure. These data can be measured on two different types of scales: interval and ratio.

1.2.2.1 Interval Scale

The interval scale is a quantitative measure. It also has an equidistant measure. But the doubling principle breaks down in this scale. The 4 marks given to an individual for his creativity do not explain that his nature is twice as good as the person with 2 marks. This is so because on this scale zero cannot be exactly located. Thus, variables measured on an interval scale have values in which differences are uniform, but ratios are not.

1.2.2.2 Ratio Scale

The data on ratio scale has a meaningful zero value and has an equidistant measure (i.e., the difference between 30 and 40 is the same as the difference between 60 and 70). For example, 60 marks obtained in a test is twice that of 30. This is so because zero exists in the ratio scale. Height is another ratio scale quantitative measure. Observations that are counted or measured are ratio data (e.g., number of goals, runs, height, and weight).

1.3 IMPORTANT DEFINITIONS

1.3.1 Variable

A variable is a phenomenon that changes from time to time, place to place, and individual to individual. It can be numeric or attribute. Numeric variable can further be classified into discrete and continuous. Discrete variable is a numeric variable that assumes value from a limited set of numbers and is always represented in whole number. Examples of such variables are number of goals, runs scored in cricket, scores in basketball match, etc. Continuous variable is also a numeric variable, but it can take any value within a range and is usually represented in fraction. Examples of such variables are height, weight, and timings.

On the other hand, an attribute is a qualitative variable that takes sub-values of a variable, such as "male" and "female," "student" and "teacher," etc. An attribute is said to be mutually exclusive if its sub-values do not occur at the same time. For instance, gender is a mutually exclusive variable because it can take value either "male" or "female" but not both. Similarly in a survey, a person can choose only one option from a list of alternatives (as opposed to selecting as many that might apply).

1.3.1.1 Independent Variable

An independent variable can be defined as the one that can be manipulated by a researcher. In planning a research experiment to see the effect of different intensities of exercise on the performance, exercise intensity is an independent variable because the researcher is free to manipulate it.

1.3.1.2 Dependent Variable

A variable is said to be dependent if it changes as a result of the change in the independent variable. In the previous example, performance is a dependent variable because it is affected by the change in exercise intensity. In fact dependent variable can be defined as the variable of interest. In creating the graph, the dependent variable is taken along the Y-axis, whereas the independent variable is plotted on the X-axis.

1.3.1.3 Extraneous Variable

Any additional variable that may provide alternative explanation or create some doubt on the conclusions in an experimental study is known as extraneous variable. If the effect of three different teaching methods on the performance is to be compared, IQ of the subjects may be considered as an extraneous variable as it might affect the final outcomes in the experiment, if IQ of all the groups are not equal...