With a nondirectional alternative hypothesis, you are predicting some type of difference, but you are not predicting the specific nature, or direction, of the difference. In some situations it might be appropriate to use a directional alternative hypothesis.
With the type of study described above, a directional alternative hypothesis also known as a one-sided or one-tailed alternative hypothesis not only predicts that there will be a difference, but also makes a specific prediction about which population will display the higher mean.
For example, in the present study, previous research might lead you to predict that the population of high goal-difficulty employees will sell more insurance, on the average, than the population of low goal-difficulty employees.
If this were the case, you might state the following directional alternative hypothesis: Statistical alternative hypothesis: In the population, mean amount of insurance sold by the high goal-difficulty group is greater than the mean amount of insurance sold by the low goal-difficulty group. Choosing directional versus nondirectional tests.
Which type of alternative hypothesis should you use in your research? Most statistics textbooks recommend using a nondirectional, or two-sided, alternative hypothesis, in most cases. The problem with the directional hypothesis is that if your obtained sample means are in the opposite direction of the direction that you predict, it can cause you to fail to reject the null hypothesis even when there are very large differences between the sample means. For example, assume that you state the directional alternative hypothesis presented above i.
If Group 2 displays the higher mean, then you might not reject the null hypothesis, no matter how great that difference might be. This presents a problem because the finding that Group 2 scored higher than Group 1 may be of great interest to other researchers particularly because it is not what many would have expected. This is why, in many situations, nondirectional tests are preferred over directional tests. Summary In summary, research projects often begin with a statement of a research hypothesis.
This allows you to develop a specific, testable statistical null hypothesis and an alternative hypothesis. Assuming the means are in the predicted direction, this type of result provides some support for your initial research hypothesis.
This type of result fails to provide support for your initial research hypothesis. Data, Variables, Values, and Observations Defining the Instrument, Gathering Data, Analyzing Data, and Drawing Conclusions With the null hypothesis stated, you can now test it by conducting a study in which you gather and analyze relevant data.
Different types of instruments can be used to obtain different types of data. For example, you might use a questionnaire to assess goal difficulty, but rely on company records for measures of insurance sold. Once the data are gathered, each agent will have one score indicating the difficulty of his or her goals, and a second score indicating the amount of insurance he or she has sold. You would then analyze the data to see if the agents with the more difficult goals did, in fact, sell more insurance.
If so, the study results would lend some support to your research hypothesis; if not, the results would fail to provide support. In either case, you would be 22 Step-by-Step Basic Statistics Using SAS: Student Guide able to draw conclusions regarding the tenability of your hypotheses, and would have made some progress toward answering your research question.
The information learned in the current study might stimulate new questions or new hypotheses for subsequent studies, and the cycle would repeat. For example, if you obtained support for your hypothesis with a correlational study, you might choose to follow it up with a study using a different research method, perhaps an experimental study the difference between these methods will be described below.
Over time, a body of research evidence would accumulate, and researchers would be able to review this body to draw general conclusions about the determinants of insurance sales. Variables, Values, and Observations Definitions. When discussing data, one often speaks in terms of variables, values, and observations.
Further complicating matters is the fact that researchers make distinctions between different types of variables such as quantitative variables versus classification variables.
This section discusses the distinctions between these terms. For the type of research discussed in this book, a variable refers to some specific characteristic of a subject that can assume one or more different values.
Subject age was a third variable, while subject sex male versus female was yet another. A value, on the other hand, refers to either a particular subject's relative standing on a quantitative variable, or a subject's classification within a classification variable. Subject age is another quantitative variable that can assume a wide variety of values. In the sample studied, these values ranged from a low of 22 years to a high of 64 years.
You can see that, in both of these examples, a particular value is a type of score that indicates where the subject stands on the variable. With quantitative variables, numbers typically serve as values.
A different type of variable is a classification variable or, alternatively, qualitative variable or categorical variable. With classification variables, different values represent different groups to which the subject might belong. These variables are classification variables and not quantitative variables because the values only represent membership in a singular, specific group—— membership that cannot be represented meaningfully with a numeric value.
In discussing data, researchers often make references to observational units, that can be defined as the individual subjects or other objects that serve as the source of the data.
Within the behavioral sciences and education, an individual person usually serves as the observational unit under study although it is also possible to use some other entity, such as an individual school or organization, as the observational unit. In this text, the individual person is used as the observational unit in most examples. An example. For a more concrete illustration of the concepts discussed so far, consider the data set displayed in Table 2.
Information about a particular observation subject is displayed as a row running horizontally from left to right across the table. The remaining five columns report information about the five research variables that are being studied. In this example, each participant has a score on a item questionnaire about the difficulty of his or her work goals.
Depending on how they respond to the questionnaire, subjects receive a score ranging from a low of zero meaning that the subject views the work goals as extremely easy to a high of meaning that the goals are viewed as extremely difficult. A rank of 1 represents the most effective agent, and a rank of 6 represents the least effective.
Table 2. One of the variables was a classification variable sex , while the remainder were quantitative variables. The numbers or letters that appear within a particular column represent some of the values that could be assumed by that variable.
Classifying Variables According to Their Scales of Measurement Introduction One of the most important schemes for classifying a variable involves its scale of measurement. Researchers generally discuss four different scales of measurement: nominal, ordinal, interval, and ratio. Before analyzing a data set, it is important to determine which scales of measurement were used because certain types of statistical procedures require specific scales of measurement.
For example, a one-way analysis of variance generally requires that the dependent variable be an interval-level or ratio-level variable; the chi-square test of independence allows you to analyze nominal-level variables; other statistics make other assumptions about the scale of measurement used with the variables that are being studied.
Chapter 2: Terms and Concepts Used in This Guide 25 Nominal Scales A nominal scale is a classification system that places people, objects, or other entities into mutually exclusive categories. A variable that is measured using a nominal scale is a classification variable: It simply indicates the name of the group to which each subject belongs. The examples of classification variables provided earlier e.
With the remaining three scales of measurement, however, some quantitative information is provided. Ordinal Scales Values on an ordinal scale represent the rank order of the subjects with respect to the variable that is being assessed.
For example, Table 2. However, an ordinal scale has a serious limitation in that equal differences in scale values do not necessarily have equal quantitative meaning. Now notice that Jim was ranked 5 while Mack was ranked 6. Putting the two together, we can see that the difference in ranking between Walt and Bob is equal to the difference in ranking between Jim and Mack.
But does this mean that the difference in overall effectiveness between Walt and Bob is equal to the difference in overall effectiveness between Jim and Mack? Not necessarily. These rankings tell us very little about the quantitative differences between the subjects with regard to the underlying construct effectiveness, in this case. An ordinal scale simply provides a rank order of who is better than whom. Interval Scales With an interval scale, equal differences between scale values do have equal quantitative meaning.
For this reason, you can see that the interval scale provides more quantitative information than the ordinal scale. A good example of an interval scale is the Fahrenheit scale used to measure temperature. With the Fahrenheit scale, the difference between 70 degrees and 75 degrees is equal to the difference between 80 degrees and 85 degrees: the units of measurement are equal throughout the full range of the scale. However, the interval scale also has an important limitation: it does not have a true zero point.
A true zero point means that a value of zero on the scale represent zero quantity of the construct being assessed. It should be obvious that the Fahrenheit scale does not have a true zero point. When the thermometer reads zero degrees, that does not mean that there is absolutely no heat present in the environment——it is still possible for the temperature to go lower into the negative numbers. For example, in the preceding study involving insurance agents, you would probably assume that scores from the goal difficulty questionnaire constitute an interval-level scale; that is, you would likely assume that the difference between a score of 50 and 60 is approximately equal to the difference between a score of 70 and Many researchers would also assume that scores from an instrument such as an intelligence test are also measured at the interval level of measurement.
On the other hand, some researchers are skeptical that instruments such as these have true equal-interval properties, and prefer to refer to them as quasi-interval scales.
Disagreements concerning the level of measurement achieved with such paper-and-pencil instruments continues to be a controversial topic within many disciplines. In any case, it is clear that there is no true zero point with either of the preceding instruments: a score of zero on the goal difficulty scale does not indicate the complete absence of goal difficulty, and a score of zero on an intelligence test does not indicate the complete absence of intelligence.
A true zero point can be found only with variables measured on a ratio scale. Chapter 2: Terms and Concepts Used in This Guide 27 Ratio Scales Ratio scales are similar to interval scales in that equal differences between scale values do have equal quantitative meaning.
However, ratio scales also have a true zero point, which gives them an additional property: with ratio scales, it is possible to make meaningful statements about the ratios between scale values. For example, the system of inches used with a common ruler is an example of a ratio scale.
With this scale, it is possible to make meaningful statements about ratios. It is appropriate to say that an object four inches long is twice as long as an object two inches long. Age, as measured in years, is also on a ratio scale: a year-old house is twice as old as a 5-year-old house. Notice that it is not possible to make these statements about ratios with the interval-level variables discussed above. One would not say that a person with an IQ of is twice as intelligent as a person with an IQ of 80, as there is no true zero point with that scale.
Although ratio-level scales are most commonly used for reporting the physical properties of objects e. Classifying Variables According to the Number of Values They Display Overview The preceding section showed that variables can be classified according to their scale of measurement.
Sometimes is also useful to classify variables according to the number of values they display. There might be any number of approaches for doing this, but this guide uses a simple division of variables into three groups according to the number of possible values: dichotomous variables, limited-value variables, and multi-value variables. Dichotomous Variables A dichotomous variable is a variable that assumes just two values. These variables are sometimes called binary variables.
You begin with 20 rats, and randomly assign them to two groups. Ten rats are assigned to the mg group they receive mg of ginkgo , and the other ten rats are assigned to the 0 mg group they receive no ginkgo. You begin with rats, and randomly assign them to four groups: Twenty-five rats are assigned to the mg group, 25 rats are assigned to the mg group, 25 rats are assigned to the 50 mg group, and 25 rats are assigned to the 0 mg group. With the Smith Anxiety Test, scores values may range from 0—99, with higher scores indicating greater anxiety.
One subject received a score of 5. Two subjects received a score of Five subjects received a score of Seven subjects received a score of Eight subjects received a score of Nine subjects received a score of Six subjects received a score of One subject received a score of Other subjects received yet other scores.
Clearly, scores on the Smith Anxiety Test constitute a multi-value variable in your sample because your subjects displayed more than six values on this variable. First, you teach each rat that, if it can correctly find its way through a maze, it will be rewarded with food at the end. You then allow each rat to try to find its way through a series of mazes. Each rat is allowed 30 trials——30 opportunities to get through a maze. Your measure of learning, therefore, is the number of mazes that each rat correctly negotiates.
This score can range from zero if the rat is not successful on any of the trials , to 30 if the rat is successful on all of the trials. A rat also can score anywhere in between these extremes.
Two rats displayed three successful trials. Three rats displayed eight successful trials. Four rats displayed 10 successful trials. Five rats displayed 14 successful trials. Six rats displayed 15 successful trials. Six rats displayed 19 successful trials.
Two rats displayed 21 successful trials. One rat displayed 27 successful trials. One rat displayed 28 successful trials. Other rats displayed yet other scores. Much research can be described as being either nonexperimental or experimental in nature. In nonexperimental research also called 30 Step-by-Step Basic Statistics Using SAS: Student Guide correlational, nonmanipulative, or observational research , the researcher simply studies the naturally-occurring relationship between two or more naturally-occurring variables.
A naturally-occurring variable is a variable that is not manipulated or controlled by the researcher; it is simply measured as it normally exists. The insurance study described previously is a good example of nonexperimental research, in that you simply measured two naturally-occurring variables goal difficulty and amount of insurance sold to determine whether they were related.
If, in a different study, you investigated the relationship between IQ and college grade point average GPA , this would also be an example of nonexperimental research. Criterion versus predictor variables. With nonexperimental designs, researchers often refer to criterion variables and predictor variables. A criterion variable is an outcome variable that can be predicted from one or more predictor variables.
The criterion variable is often the main focus of the study in that it is the outcome variable mentioned in the statement of the research problem. With our insurance example, the criterion variable is the amount of insurance sold. The predictor variable, on the other hand, is the variable that is used to predict values on the criterion. In some studies, you might even believe that the predictor variable has a causal effect on the criterion.
You do not necessarily have to believe that there is a causal relationship between two variables to conduct a study such as this, however; you might simply be interested in determining whether it is possible to predict one variable from the other.
Cause-and-effect relationships. It should be noted here that nonexperimental research that investigates the relationship between just two variables generally provides very weak evidence concerning cause-and-effect relationships. The reasons for this can be seen by reviewing our study on insurance sales. If the psychologist conducts this study and finds that the agents with the more difficult goals also tend to sell more insurance, does that mean that having difficult goals caused them to sell more insurance?
It can also be argued that selling a lot of insurance increases the agents' self-confidence, and that this causes them to set higher work goals for themselves. Under this second scenario, it was actually the insurance sales that had a causal effect on goal difficulty. As this example shows, with nonexperimental research it is often possible to obtain a single finding that is consistent with a number of different, contradictory causal explanations.
To obtain stronger evidence of cause and effect, researchers generally either analyze the relationships among a larger number of variables using sophisticated statistical procedures that are beyond the scope of this text such as structural equation modeling , or drop the nonexperimental approach entirely and instead use experimental research methods.
The nature of experimental research is discussed in the following section. To illustrate these concepts, let's describe a possible experiment in which you test the hypothesis that goal difficulty positively affects insurance sales. First you identify a group of agents who will serve as subjects. Assume that this is a relatively difficult goal. They have been told to make just 5 cold calls to potential policy holders per week.
To the extent possible, you see to it that agents in both groups are treated similarly with respect to everything except for the difficulty of the goals that are set for them. After one year, you determine how much new insurance each agent has sold that year. Independent versus dependent variables. It is possible to use some of the terminology associated with nonexperimental research when discussing this experiment. For example, it would be appropriate to continue to refer to the amount of insurance sold as being a criterion variable because this is the outcome variable of central interest.
You also could continue to refer to goal difficulty as the predictor variable because you believe that this variable will predict sales to some extent. Notice that goal difficulty is now a somewhat different variable, however.
In the nonexperimental study, goal difficulty was a naturally-occurring variable that could take on a wide variety of values whatever score the subject received on the goal difficulty questionnaire.
In the present experiment, however, goal difficulty is a manipulated variable, which means that you as the researcher determined what value of the variable would be assigned to each subject. In the experiment, the goal difficulty variable could assume only one of two values: Subjects were either in the difficult goal group or the easy goal group.
Therefore, goal difficulty is now a classification variable that codes group membership. Although it is acceptable to speak of predictor and criterion variables within the context of experimental research, it is more common to speak in terms of independent variables and dependent variables.
The independent variable is that variable whose values or levels are 32 Step-by-Step Basic Statistics Using SAS: Student Guide selected by the experimenter to determine what effect the independent variable has on the dependent variable. The independent variable is the experimental counterpart to a predictor variable. A dependent variable, on the other hand, is some aspect of the subject's behavior that is assessed to determine whether it has been affected by the independent variable.
The dependent variable is the experimental counterpart to a criterion variable. In the present example experiment, goal difficulty is the independent variable, and the amount of insurance sold is the dependent variable.
Levels of the independent variable. Researchers often speak in terms of the different levels of the independent variable. These levels are also referred to as experimental conditions or treatment conditions, and correspond to the different groups to which a subject might be assigned. The present example included two experimental conditions: a difficult goal condition, and an easy goal condition. With respect to the independent variable, it is common to speak of the experimental group versus the control group.
Generally speaking, the experimental group is the group that receives the experimental treatment of interest, while the control group is an equivalent group of subjects that does not receive this treatment. The simplest type of experiment consists of one experimental group and one control group. For example, the present study could have been redesigned so that it simply consisted of an experimental group that was assigned the goal of making 25 cold calls the difficult goal condition , as well as a control group in which no goals were assigned the no-goal condition.
Obviously, it is possible to expand the study by creating more than one experimental group. This could be accomplished in the present case by assigning one experimental group the difficult goal of 25 cold calls and the second experimental group the easy goal of 5 cold calls. The control group could still be assigned zero goals. Using Type-of-Variable Figures to Represent Dependent and Independent Variables Overview Many studies in the social sciences and education are designed to investigate the relationship between just two variables.
In an experiment, researchers generally refer to these as the independent and dependent variables; in a nonexperimental study, researchers often call them the predictor and criterion variables. Chapter 2: Terms and Concepts Used in This Guide 33 Some chapters in this guide will describe studies in which a researcher investigates the relationship between predictor and criterion variables.
To help you better visualize the nature of the variables being analyzed, most of these chapters will provide a type-ofvariable figure: a figure that graphically illustrates the number of values that are assumed by the two variables in the study. This section begins by presenting the symbols that will represent three types of variables: dichotomous variables, limited-value variables, and multi-value variables.
It then provides a few examples of the type-of-variable figures that you will see in subsequent chapters of this book. Figures to Represent Types of Variables Dichotomous variables.
A dichotomous variable is one that assumes just two values. Limited-value variables. A limited-value variable is one that assumes only two to six values.
Multi-value variables. A multi-value variable is one that assumes more than six values. It is possible to construct a type-of-variable figure that illustrates the nature of the dependent variable, as well as the nature of the independent variable, in a single figure. The research hypothesis. For example, earlier this chapter developed the research hypothesis that goal difficulty will have a positive causal effect on the amount of insurance sold by insurance agents.
This hypothesis was illustrated by the causal figure presented in Figure 2. That figure is again reproduced here as Figure 2.
Notice that, in this figure, the dependent variable amount of insurance sold appears on the left, and the independent variable goal difficulty appears on the right. Predicted causal relationship between goal difficulty the independent variable and amount of insurance sold the dependent variable. An experiment with two conditions. In this example, you conduct a simple experiment to investigate this research hypothesis.
You begin with insurance agents, and randomly assign each agent to either an experimental group or a control group. After one year, you measure your dependent variable: The amount of insurance in dollars sold by each agent. As a group, they displayed far more than six values on this dependent variable.
You knew this, because the agents displayed more than six values on this variable. You knew this, because this independent variable consisted of just two values conditions : a difficult-goal condition and an easy-goal condition.
Because the dependent variable is on the left and the independent variable is on the right, the preceding type-of-variable figure is similar to Figure 2. In that figure, the dependent variable was also on the left, and the independent variable was also on the right The preceding type-of-variable figure could be used to illustrate any experiment in which the dependent variable was a multi-value variable and the independent variable was a dichotomous variable.
A warning about statistical assumptions. Please note that, when you are deciding whether it is appropriate to analyze a data set with a t test, it is not sufficient to simply verify that the dependent variable is a multi-value variable and that the independent variable is a dichotomous variable. There are many statistical assumptions that must be satisfied for a t test to be appropriate, and those assumptions will not be discussed in this chapter they will be discussed in the chapters on t tests.
The type-of-variable figure was presented above to help you visualize the type of situation in which a t test is often performed.
Each chapter of this guide that discusses an inferential statistical procedure such as a t test also will describe the assumptions that must be met in order for the test to be valid. An experiment with three conditions. Assume that everything else about the study remains the same. That is, you use the same dependent variable, the number of values observed on the dependent variable still exceed six, and so forth. This means that the independent variable is now a limited-value variable, not a dichotomous variable.
The preceding figure could be used to illustrate any experiment in which the dependent variable was a multi-value variable, and the independent variable was a limited-value variable.
A correlational study. This time you are interested in the same research hypothesis, but you are doing a nonexperimental study rather than an experiment. In this study, you will not manipulate an independent variable.
Instead, you will simply measure two naturally occurring variables and will determine whether they are correlated in a sample of insurance agents. With this scale, scores can range from 0 to 99, with higher scores representing more difficult goals.
When you analyze the data, you find that this variable displays more than six values in this sample i. For each agent, you review records to determine how much insurance the agent has sold during the previous year.
Assume that this variable also displays more than six observed values in your sample. By analyzing your data, you want to determine whether there is a significant correlation between goal difficulty and the amount of insurance sold. You hope to find that agents who had high scores on goal difficulty also tended to have high scores on insurance sold. Because this is nonexperimental research, it is not appropriate to speak in terms of an independent variable and a dependent variable.
When preparing a type-ofvariable figures for this type of study, the criterion variable should appear to the left of the equals sign, and the predictor variable should appear to the right. You knew that it was a multi-value variable, because it displayed more than six values in your sample. The preceding figure could be used to illustrate any correlational study in which the criterion variable and predictor variable were both multi-value variables. The Three Types of SAS Files Overview The purpose of this section is to provide a very general overview of the procedure that you will follow when you submit a SAS program and then interpret the results.
To do this, the current section will present a short SAS program and briefly describe the output that it creates. The differences between these three types of files are discussed next.
These statements provide the SAS System with the data to be analyzed, tell SAS about the nature of the data, and indicate which statistical analyses should be performed on the data. Some fictitious data. This section will illustrate a simple SAS program that analyzes some fictitious data.
Suppose that you have administered two tests Test 1 and Test 2 to a group of eight people. Scores on a particular test can range from 0 to 9. Each vertical column running from the top to the bottom provides information about a different variable.
The headings in Table 2. In contrast, each horizontal row in the table running from left to right provides information about a different subject. The rows for the remaining subjects can be interpreted in the same way. Suppose that you now want to analyze subject scores on the two tests. Specifically, you want to compute the means and some other descriptive statistics for Test 1 and Test 2. Following is a complete SAS program that enters the data presented in Table 2.
It also computes means and some other descriptive statistics for Test 1 and Test 2. We will then be able to use these line numbers to refer to specific statements.
This is a global statement that can be used to modify how the SAS System operates. You use this statement to start the DATA step explained below and assign a name to the data set that you are creating.
You use this statement to assign names to the variables that SAS will work with. This statement tells SAS that the data lines will begin with the next line of the program. You can see that these data lines were taken directly from Table 2. There are eight data lines because there were eight subjects. This null statement tells SAS that the data lines have ended.
It tells SAS to compute means and other descriptive statistics for all numeric variables in the data set. You use this statement to assign a title, or heading, that will appear on each page of output.
Subsequent chapters will discuss the use of the preceding statements in much more detail. What is the single most common programming error? For new SAS users, the most common programming error usually involves omitting a required semicolon ;. When you obtain an error in running a SAS program, one of the first things that you should do is inspect the program for missing semicolons. There is another, more fundamental way, to divide a SAS program into its constituent components.
For example, the PROC step might request that correlations be computed for all pairs of numeric variables, or might request that a t test be performed. In the preceding example, the PROC step requested that means be computed. What text editor will I use to write my SAS program? An editor is a computer application that allows you to create lines of text, such as the lines that constitute a SAS program. If you are working on a mainframe or mid range computer system, you might have a variety of editors that can be used to write your SAS programs; just ask the staff at your computer facility.
The SAS windowing environment is an integrated application that allows users to create and edit SAS programs, submit them for interactive analysis, view the results on their screens, manage files, and perform other activities. This application is available at most locations where the SAS System is installed including personal computers. Chapter 3 of this guide provides a tutorial that shows you how to use the SAS windowing environment. After submitting the SAS program.
Once the preceding program has been submitted for analysis, SAS will create two types of files reporting the results of the analysis. The following sections explain the purpose of these files. It is a summary of notes and messages generated by SAS as your program executes. These notes and messages will help you verify that your SAS program ran correctly. D1 has 8 observations and 2 variables. Log file created by the current SAS program.
Notice that the statements constituting the SAS program have assigned line numbers, which are reproduced in the SAS log. The data lines are not normally reproduced as part of the SAS log unless they are specifically requested. This note indicates that the data set that you created named D1 contains 8 observations and 2 variables. You would normally check this note to verify that the data set contains all of the variables that you intended to input in this case 2 , and that it contains data from all of your subjects in this case 8.
So far, everything appears to be correct. Often, these error messages provide you with some help in determining what was wrong with the program. For example, a message can indicate that SAS was expecting a program statement that was not included.
Once the error or errors have been identified, you must revise the original SAS program and resubmit it for analysis. If the log indicates that the program ran correctly, you are free to review the results of the analyses in the SAS output file.
Very often you will submit a SAS program and, after a few seconds, the SAS output window will appear on your computer screen. Some users mistakenly assume that this means that their program ran without errors. But this is not necessarily the case. Very often some parts of your program will run correctly, but other parts will have errors. Chapter 3 will lead you through these steps. Because the program above requested the MEANS procedure, the output file that was produced by this program will contain means, standard deviations, and some other descriptive statistics for the two variables.
Output 2. Numbers such as and have been added to the output to more easily identify specific sections. Later, this guide will show you how to insert your name in the TITLE1 statement, so that your name will appear at the top of each of your output pages. For Test 2, the corresponding figures were 3. Once you have obtained this output file from your analysis, you can review it on your computer monitor, or print it out at a printer.
Chapter 3 will show you how to interpret your output. Conclusion This chapter has introduced you or reintroduced you to the terminology that is used by researchers in the behavioral sciences and education. With this foundation, you are now ready to learn about performing data analyses with SAS. The preceding section indicated that you must use some type of text editor to write SAS programs. For most users, it is advantageous to use the SAS windowing environment for this purpose.
With the SAS windowing environment, you can write and submit SAS programs, view the results on your monitor, print the results, and save your SAS programs on a diskette——all from within one application. Chapter 3 provides a hands-on tutorial that shows you how to perform these activities within the SAS windowing environment.
The tutorial in this chapter is based on Version 8 of SAS. If you are using Version 7 of SAS, you can still use the tutorial presented here with some minor adjustments , because the interfaces for Version 7 and Version 8 are very similar. However, if you are using Version 6 of SAS, the interface that you are using is substantially different from the Version 8 interface. The majority of this chapter consists of a tutorial that is divided into four parts.
Part III takes you through the steps involved in debugging a program with an error. Finally, Part IV gives you the opportunity to practice what you have learned. In addition, two short sections at the end of the chapter summarize the steps that are involved in frequently performed activities, and show you how to use the OPTIONS statement to control the size of your output page.
Friend Reviews. To see what your friends thought of this book, please sign up. Lists with This Book. This book is not yet featured on Listopia. Add this book to your favorite list ». Community Reviews.
Showing Average rating 3. Rating details. More filters. Sort order. Aug 27, Tanya rated it it was ok Shelves: it , school , sas , statistics. Easier to use the Internet. Rishabh rated it liked it Jan 31, Vidhi Gupta rated it liked it Feb 23, Brittany rated it it was amazing Mar 17, Diana Douglas added it Jan 01, Sai marked it as to-read Jul 16, Getting comfortable with interface is a matter of a few minutes and companies interview candidates based on programming knowledge not on interface.
SAS Statistics may take a longer period of time if you don't have prior knowledge of Statistics. SAS Statistical procedures are straight-forward but how they are used in real-world case studies requires professional experience or you can try to solve kaggle case studies using SAS.
You can follow the links above and put your question related to the topic in the comment section. SAS also has official community for questions and answers. It includes many base and advanced tutorials which would help you to get started with SAS and you will acquire knowledge of data exploration and manipulation, predictive modeling using SAS along with some scenario based examples for practice.
Table of Contents. What is SAS? SAS Statistical analysis system is one of the most popular software for data analysis. It is widely used for various purposes such as data management, data mining, report writing, statistical analysis, business modeling, applications development and data warehousing.
Knowing SAS is an asset in many job markets.
0コメント