Business Research Methods An overview

108

Nature of multivariate analysis

Data does not always come with a single response nor does it always have a response. A data set may consist simply n measurements on p variables. For example, a doctor might record patients height, weight, blood pressure and pulse. We could envision situations where any one of these variables is the response and the others are predictors. Or we could have a situation where we just wanted to examine similarities (and differences) of patients As the name indicates, multivariate analysis comprises a set of techniques dedicated to the analysis of data sets with more than one variable. Several of these techniques were developed recently in part because they require the computational capabilities of modern computers.

Statistical Techniques
Univariate Multivariate

UNIVARIATE:- When there are many observations about a variable and if the variable has to be individually analyzed, univariate technique is to be applied. Univariate data are of two types: Non-metric data:- Nominal or ordinal scale Metric data:- Interval or ratio scale

108

Business Research Methods An overview

109

Non-Metric Data
Indipendent 1. Rank sum test 2. Kruskal Wallis test 3. Mann Whitney test Dependent 1. Chi-Square test 2. Kolmegrov-Smirnov 3. Runs 4. Wilcexon 5. Mcnemar 6. Cochran q-test

Multivariate Analysis
It helps to analyze more than two variables simultaneously on a sample of observations. It takes into account various relationships among the variables. They are largely empirical and deal with reality. They analyze complex data. It helps in decision making and helps to derive the probability statements.

Multivariate technique

109

Business Research Methods An overview

110

MULTIPLE REGRESSION
The general purpose of multiple regression (the term was first used by Pearson, 1908) is to learn more about the relationship between several independent or predictor variables and a dependent or criterion variable. For example, a real estate agent might record for each listing the size of the house (in square feet), the number of bedrooms, the average income in the respective neighborhood according to census data, and a subjective rating of appeal of the house. Once this information has been compiled for various houses it would be interesting to see whether and how these measures relate to the price for which a house is sold. For example, one might learn that the number of bedrooms is a better predictor of the price for which a house sells in a particular neighborhood than how "pretty" the house is (subjective rating). One may also detect "outliers," that is, houses that should really sell for more, given their location and characteristics. Personnel professionals customarily use multiple regression procedures to determine equitable compensation. One can determine a number of factors or dimensions such as "amount of responsibility" (Resp) or "number of people to supervise" (No_Super) that one believes to contribute to the value of a job. The personnel analyst then usually conducts a salary survey among comparable companies in the market, recording the salaries and respective characteristics (i.e., values on dimensions) for different positions. This information can be used in a multiple regression analysis to build a regression equation of the form: Salary = .5*Resp + .8*No_Super Once this so-called regression line has been determined, the analyst can now easily construct a graph of the expected (predicted) salaries and the actual salaries of job incumbents in his or her company. Thus, the analyst is able to determine which position is underpaid (below the regression line) or overpaid (above the regression line), or paid equitably. In the social and natural sciences multiple regression procedures are very widely used in research. In general, multiple regression allows the researcher to ask (and hopefully answer) the general question "what is the best predictor of ...". For example, educational researchers might want to learn what are the best predictors of success in high-school. Psychologists may want to determine which personality variable best predicts social adjustment. Sociologists may want to find out which of the multiple social indicators best predict whether or not a new immigrant group will adapt and be absorbed into society When there are two or more than two independent variables, the equation describing such a relationship is called multiple regression equation. Simple linear equation:Y= a + bX X =independent variable Y= dependent variable a=intercept
110

Business Research Methods An overview

111

b=slope Multiple regression equation

Y = a+ b1X1 + b2X2 + b3X3

Here X1,X2 and X3 are independent variables. a= intercept b1, b2, b3 are slopes of X1, X2 & X3 Y= dependent variable For example:How the annual sales revenue of a company is dependent? It may be dependent on sales force, annual advertising expenditure efficient marketing etc. which are the independent variables. A regression model can be constructed to forecast the annual sales revenue. REGRESSION MODEL In the multivariate case, when there is more than one independent variable, the regression line cannot be visualized in the two dimensional space, but can be computed just as easily. For example, if in addition to IQ we had additional predictors of achievement (e.g., Motivation, Self- discipline) we could construct a linear equation containing all those variables. In general then, multiple regression procedures will estimate a linear equation of the form: Y = a + b1*X1 + b2*X2 + ... + bp*Xp So Y= a + b1X1+ b2X2

Using the normal equations and solve the constants a, b1, b2 by the three equations: Y = na + bX1+ b2X2

YX1 = aX1 + b1(X1)+ b2X1X2 YX2 = aX2 + b1X1X2 + b2(X2) n= total no. of observations

The no. of normal equations depends on the no. of independent variables. If there are two independent variables, then there will be three equations and if there are three independent variables, then there will be four equations.

111

Business Research Methods An overview

112

If the independent variable has high degree of correlation, then the regression coefficients a,b1,b2 becomes less reliable. This problem has been called multi-co linearity. If variable X1 is correlated with X2, then it distorts the value of regression coefficients. Prediction of dependent variable can be done even if multi-collinearity is present. The multi-collinearity should be minimum in predicting the dependent variable.

DISCRIMINANT ANALYSIS
Discriminant analysis is a technique for classifying a set of observations into predefined classes. The purpose is to determine the class of an observation based on a set of variables known as predictors or input variables. The model is built based on a set of observations for which the classes are known. This set of observations is sometimes referred to as the training set. Based on the training set , the technique constructs a set of linear functions of the predictors, known as discriminant functions, such that L = b1x1 + b2x2 + + bnxn + c , where the b's are discriminant coefficients, the x's are the input variables or predictors and c is a constant. These discriminant functions are used to predict the class of a new observation with unknown class. For a k class problem k discriminant functions are constructed. Given a new observation, all the k discriminant functions are evaluated and the observation is assigned to class i if the ith discriminant function has the highest value. 1. Individuals or objects are classified into groups based on a set of independent variables. 2. This is a type of regression equation, where the dependent variable represent a group. 3. This discriminate function discriminates between members of group and says the member belong to which group. 4. The dependent variable is non-metric (nominal or ordinal). 5. The independent variable is metric (interval or ratio). The Simplest Case The simplest case of discriminant function analysis is the prediction of dichotomous group membership based on a single variable. An example of the simplest case is the prediction of successful completion of a graduate program based on the GRE verbal score. In this case, since the prediction model includes only a single variable, it gives little insight into how variables interact with each other in prediction. Thus prediction of group membership will be the major focus of the next section of this chapter. With respect to the data file and purpose of analysis, this simplest case is identical to the case of linear regression with dichotomous dependent variables. As discussed previously, data of this type may be represented in any number of different forms: scatterplots, tables of means and standard deviations, and overlapping frequency polygons. Because overlapping frequency polygons have such an intuitive appeal, they will be used to describe how discriminant function analysis works.

112

Business Research Methods An overview

113

Example three categories namely clerk, manager and custodies are studies for a liking of a brand of cigarette. The buying capacity of the cigarette varies among the three group members. It is very much influence by there age group, income, work stress, family etc. Now we are trying to study which category of the specific age group and income likes this particular brand of cigarette. Actually we are assigning a score individuals liking of the cigarette. It means we are giving a weighted average to individuals liking of the cigarette. When calculating we calculate the significance through Wilks Lambda and low the value indicates the high significance. When calculating we calculate the significance through Wilks Lambda and low the value indicates the high significance.

Further go to this link http://www.psychstat.missouristate.edu/multibook/mlt03.htm

MANOVA
Multivariate analysis of variance (MANOVA) is simply an ANOVA with several dependent variables. Multivariate analysis of variance (MANOVA) is a generalized form of analysis of variance (ANOVA) methods to cover cases where there is more than one (correlated) dependent variable and where the dependent variables cannot simply be combined. For example, we may conduct a study where we try two different textbooks, and we are interested in the students' improvements in math and physics. In that case, improvements in math and physics are the two dependent variables, and our hypothesis is that both together are affected by the difference in textbooks. A multivariate analysis of variance (MANOVA) could be used to test this hypothesis. Instead of a univariate F value, we would obtain a multivariate F value (Wilks' ?) based on a comparison of the error variance/covariance matrix and the effect variance/covariance matrix. Although we only mention Wilks' ? here, there are other statistics that may be used, including Hotelling's trace and Pillai's criterion. The "covariance" here is included because the two measures are probably correlated and we must take this correlation into account when performing the significance test. Testing the multiple dependent variables is accomplished by creating new dependent variables that maximize group differences. These artificial dependent variables are linear combinations of the measured dependent variables. Research Questions The main objective in using MANOVA is to determine if the response variables (student improvement in the example mentioned above), are altered by the observers manipulation of the independent variables. Therefore, there are several types of research questions that may be answered by using MANOVA:

113

Business Research Methods An overview

114

1) What are the main effects of the independent variables? 2) What are the interactions among the independent variables? 3) What is the importance of the dependent variables? 4) What is the strength of association between dependent variables? 5) What are the effects of covariates? How may they be utilized?

In MANOVA instead of single variable, a set of variables are taken. 1. The F ratio of among groups variance within groups variance is calculated. 2. Here several metric dependent variables are studied.

For example a researcher want to test a hypothesis concerning multivariate differences is group responses to experimental manipulations. A researcher studies the contamination status of fishes is a fond and assess its suitability for human consumption. He takes two ponds :One is pristine or called as control area & the other is polluted or called a test area. He analyzes the protein quality wt of the fish, length continuant level and tries to examine the variables between these two areas with respect to the above said variables and thus assess their status for human consumption. In the analysis we will obtain a multivariate F value called as Wilks Lambda. Here we compare the error variance metric and the effect variances matrix. Assumptions Normal Distribution: - The dependent variable should be normally distributed within groups. Overall, the F test is robust to non-normality, if the non-normality is caused by skewness rather than by outliers. Tests for outliers should be run before performing a MANOVA, and outliers should be transformed or removed. Linearity - MANOVA assumes that there are linear relationships among all pairs of dependent variables, all pairs of covariates, and all dependent variable-covariate pairs in each cell. Therefore, when the relationship deviates from linearity, the power of the analysis will be compromised. Homogeneity of Variances: - Homogeneity of variances assumes that the dependent variables exhibit equal levels of variance across the range of predictor variables. Remember that the error variance is computed (SS error) by adding up the sums of squares within each group. If the variances in the two groups are different from each other, then adding the two together is not appropriate, and will not yield an estimate of the common within-group

114

Business Research Methods An overview

115

variance. Homoscedasticity can be examined graphically or by means of a number of statistical tests. Homogeneity of Variances and Covariances: - In multivariate designs, with multiple dependent measures, the homogeneity of variances assumption described earlier also applies. However, since there are multiple dependent variables, it is also required that their intercorrelations (covariances) are homogeneous across the cells of the design. There are various specific tests of this assumption.

WebPages: Site Stat soft text entry on MANOVA EPA Statistical Primer Introduction to MANOVA Practical guide to MANOVA for SAS Link http://www.statsoft.com/textbook/stathome.html http://www.epa.gov/bioindicators/primer/html/manova.html http://ibgwww.colorado.edu/~carey/p7291dir/handouts/manova1.pdf http://ibgwww.colorado.edu/~carey/p7291dir/handouts/manova2.pdf

VARIABLES IN MULTI VARIATE ANALYSIS Explanatory variable & criterion variables They are nothing but independent and dependent variable. For example: X is responsible for production of Y. X explanatory or casual variable Y responded variable The set of X1, X2, X3 are a set of explanatory variables and Y1, Y2, Y3 are a set of criterion variables. In economics explanatory variables external or exogenous. Criterion internal or endogenous variable. Observable and latest variable: When independent or explanatory variables are directly observable, they are called as observable variables. Some variable which are not seen and influence the criterion variables, they are called latent variables.

115

Business Research Methods An overview

116

LISREL
Through this we construct structural and measurement model in a combined manner. STRUCTURAL MODEL Based on data we construct a structural model for eg. We take 100 gm of pulses and ground them in a mixer and we expect 95 gm as powder and 5 gm as loss . (a)100gm (b) 95 gm output (c) loss =5 gm thus output=input-loss b=c-a We form a structural equation .We presume that the variable is measured (wt. of the pulses in gm.) without any error. This assumption is not reasonable. The output is the dependent variable and during its measurement error is expected and its random measurement error does not lead to large standard errors. To deal the measurement error factor analysis is used . In the factor analysis, large number of items is reduced to a smaller number of factors. The output is weighed. The error might occur due to : 1. Spillage of the sample. 2. Weighing accuracy and precision of the machine. 3. Grinding capability and the loss. 4. External unseen factors. 5. Investigators precision and accuracy. Thus the various scales will be reduced to a common locus of control scale and it could be more reliable.

Factor analysis can be either:

Exploratory: the computer determines what the underlying factors are. Confirmatory: the researcher specifies what factor structure that he/she thinks underlies the measures and then tests. Whether the data are consistent with his/her hypothesis.

Programs such as LISREL makes it possible to combine structural equations modeling and confirmatory factor analysis. LISREL can handle a wide array of problems and models. These include : 1. Models with measurement error 2. Non recursive models 3. Manova type problems 4. Multiple group comparisons 5. Tests of constraints 6. Confirmatory factor analysis 7. Ordinal regression 8. Hierarchical linear models

116

Business Research Methods An overview

117

LISREL helps to ensure the group differences and to identify which interaction effects should be in the models.

CONJOINT ANALYSIS
Conjoint analysis is a popular marketing research technique that marketers use to determine what features a new product should have and how it should be priced. Conjoint analysis became popular because it was a far less expensive and more flexible way to address these issues than concept testing. The basics of conjoint analysis are not hard to understand. A simple example is all thats required. Suppose we want to market a new golf ball. We know from experience and from talking with golfers that there are three important product features: Average Driving Distance Average Ball Life Price We further know that there is a range of feasible alternatives for each of these features, for instance: Average Driving Distance Average Ball Life Price 275 yards 54 holes $1.25 250 yards 36 holes $1.50 225 yards 18 holes $1.75 Obviously, the markets ideal ball would be: Average Driving Distance Average Ball Life Price 275 yards 54 holes $1.25 and the ideal ball from a cost of manufacturing perspective would be: Average Driving Distance Average Ball Life Price 225 yards 18 holes $1.75 assuming that it costs less to produce a ball that travels a shorter distance and has a shorter life.Heres the basic marketing issue: Wed lose our shirts selling the first ball and the market wouldnt buy the second. The most viable product is somewhere in between, but where? Conjoint analysis lets us find out where. A traditional research project might start by considering the rankings for distance and ball life in Figure 1.

117

Business Research Methods An overview

118

This type of information doesnt tell us anything that we didnt already know about which ball to produce. Now consider the same two features taken conjointly. Figures 2a and 2b show the rankings of the 9 possible products for two buyers assuming price is the same for all combinations.

Both buyers agree on the most and least preferred ball. But as we can see from their other choices, Buyer 1 tends to trade-off ball life for distance, whereas Buyer 2 makes the opposite Trade-off. The knowledge we gain in going from Figure 1 to Figures 2a and 2b is the essence of conjoint analysis. If you understand this, you understand the power behind this technique. Next, lets figure out a set of values for driving distance and a second set for ball life for Buyer 1 so that when we add these values together for each ball they reproduce Buyer 1's rank orders. Figure 3 shows one possible scheme.

Notice that we could have picked many other sets of numbers that would have worked, so there is some arbitrariness in the magnitudes of these numbers even though their relationships to each other are fixed. Next suppose that Figure 4a represents the trade-offs Buyer 1 is willing to make between ball life and price. Starting with the values we just derived

118

Business Research Methods An overview

119

for ball life, Figure 4b shows a set of values for price that when added to those ball life reproduce the rankings for Buyer 1 in Figure 4a.

We now have in Figure 5 a complete set of values (referred to as utilities or part-worths) that capture Buyer 1's trade-offs.

Lets see how we would use this information to determine which ball to produce. Suppose we were considering one of two golf balls shown in Figure 6.

The values for Buyer 1 in Figure 5 when added together give us an estimate of his preferences. Applying these to the two golf balls were considering, we get the results in Figure 7.

119

Business Research Methods An overview

120

Wed expect buyer 1 to prefer the long-life ball over the distance ball since it has the larger total value. Its easy to see how this can be generalized to several different balls and to a representative sample of buyers. These three steps--collecting trade-offs, estimating buyer value systems, and making choice predictions- form the basics of conjoint analysis. Although trade-off matrices are useful for explaining conjoint analysis as in this example, not many researchers use them nowadays. Its easier to collect conjoint data by having respondents rank or rate concept statements or by using PC-based interviewing software that decides what questions to ask each respondent, based on his previous answers. As you may expect there is more to applying conjoint analysis than is presented here. But if you understand this example, you understand what conjoint analysis is. SO its a technique which deals with measurement of customer preferences 1. Another general example the customer preference on purchasing a computer is affected by number of factors:2. Price 3. RAM 4. Disk capacity And the no. of level of each factor are I. Price Rs 35000, 40000, & 50000 II. RAM 512MB, 1GB, 2GB III. Hard disk 80GB, 160GB, 250GB The factorial combination is made the objective is to find most appealing product profile & also the most profitable product profile based on the customer preferences. If the appealing and profitable product are same marketing become easier. If they vary a compromise may be arrived on the final product profile.

Approaches to Conjoint Analysis 1. Two factor evaluation 2. Multifactor evaluation

FACTOR ANALYSIS

120

Business Research Methods An overview

121

There might be a number of independent variable used in predicting a dependent

variable. Because of too many variables it leads to the following problems: Increased computational terms to get solution. Increased time in data collection. Too much expenditure in data collection. Presence of redundant independent variables. Difficulty in making inferences. It can be minimized/avoided by factor analysis. It is used when variables have high degree of inter correlation. For example a marketing manager of a two-wheeler company designed a questionnaire to study the customer feedback and in turn he is keen in identifying the factors of his study. He has identified six variables: Fuel efficiency (x1) Life of the two-wheeler (x2) Handling convenience (x3) Quality of original spares (x4) Breakdow rate (x5) Price (x6) He groups certain variables into factors x1,x2,x4,x5 factor 1 x6 factor 2 x3 factor 3 Factor 1 technical factor Factor2 price factor Factor 3 personal factor In future to conduct a detailed study it will be sufficient to get an opinion on the above said factors. Basic terminologies: Factor: it is an underlying dimension of several related variables. There might be one or more factors depending on the number and nature of variables involved in it. Factor loadings Li(j): it is a matrix representing the correlation between different combination of variables and factors. It represents how closely the variables are related to each one of the factors discovered. They are factor variable correlations. Li(j) is factor loading of the variable j on the factor i. Communality: hi It is the sum of squares of factor loadings of the variable on all factors. Hi of a variable= (sum of factor loading of a factor) It shows how much of each variable is accounted for the underlying factors. Eigen value: It is the sum of squares of factor loadings of all variables in a factor. It takes the relative importance of each factor in accounting for particular set of variables.

121

Business Research Methods An overview

122

Total sum of squares: when Eigen values of all factors are totaled, the resulting value is called total sum of squares. Rotation: After obtaining the factor loadings, the factor loading matrix structure has to be studied. If there is no simple structure, there are n dimensional spare of factors & they can be rotated in angles to simplify the process of interpretation of factors. If the factors are independent orthogonal rotation(90 degree between different factors after rotation). If they are correlated oblique rotation is done (not 90 degree between different factors even after rotation). Factor Scores: Though a factor is not visible like an input variable, it is still like a variable and can be used to find scores for the respondents. Initially the respondent assign scores for variables, after factor analysis, each factor assigns a score for the respondent. Significant number of factors: The main idea of factor analysis is to reduce the number of input variables to factors and even to reduce the number of factors & extract maximum information. The criteria for determining the number of factors to be retained for further study; minimum eigen value criterion: The eigen value for factor should be more than or equal to one or the factor is dropped. Scree plot criterion: Plot the factor numbers on x-axis & eigen values on y-axis. Identify the factor number at which the slope of the line changes from steep to a gradual trailing. Such a change is called scree & the point is known as scree point. Factors marked upto the scree point are taken for further study. Methods of factor analysis: Centroid method Devised by LL.Thurstone. It maximizes the absolute loadings of each factor. It is a less cumbersome method. Principal component method Developed by H.Holelting. Here we construct new variables from a given set of input variables which are called as principal components. The principal components are generally incorrelated. The first principal component has the maximum variance followed by the other.

CLUSTER ANALYSIS
Centroid method

122

Business Research Methods An overview

123

1. Centroid is the point whose coordinates are the means of all observations in the cluster 2. When two clusters are combined, a new cluster is formed 3. When new clusters are formed, the cluster centroid migrate or move Advantages of hierarchical clustering 1. It is easy to read & interpret 2. The output has a logical structure that theoretically always exists Disadvantages of hierarchical clustering 1. It is relatively unstable and unreliable 2. For example, if two objects are combined together which may be based on a small difference in distance may constrain the rest of the analysis NOTE: In hierarchical clustering, it is better to split the sample into atleast two groups and do independent clustering runs and see if similar clusters emerge in both runs Non hierarchical clustering 1. In this it permits the objects to leave one cluster and joins to another 2. The initial centroids or starting point of the clusters are called as cluster seeds 3. The cluster seeds groups objects within a prespecified distance 4. A cluster is initially selected and all objects within a prespecified threshold distance are included in the cluster Non hierarchical method 1. It is also known as iterative partitioning 2. The three common approaches are i. Sequential threshold ii. Parallel threshold iii. Optimizing procedures Sequential threshold 1. A cluster centre is selected and all objects within a prespecified threshold value are grouped 2. Again a new cluster centre is selected and process is repeated for the unclustered objects 3. Once the objects enter a cluster, they are removed from further processing Parallel threshold 1. Here several clusters are selected simultaneously 2. The objects having the threshold level closer to themselves are assigned to the nearest central cluster Optimizing 1. It modifies the previous two procedures of parallel and sequential threshold 2. It optimizes the overall criterion measure where the objects are assigned to clusters with optimum average within cluster distance Measures of similarity 1. Grouping of objects together- similarity or dissimilarity measures are needed 2. Similar objects are grouped together while others are put in separate clusters 3. Commonly used measures for cluster analysis 1. Distance measures

123

Business Research Methods An overview

124

2. Correlation coefficients 3. Association coefficients 4. The most popular measure of distance is the Squared Euelidean distance 1. In a data matrix each variable is measured in different unit the variables have to be standardized to zero mean and unit standard deviation 2. Thus standardization can remove the influence of unit of measurement and also reduce the differences between groups or variables 3. Observations with extreme values should be removed 4. Drawback of the distance measure is variable with both large size differences and 5. standard deviation can essentially swamp the effects of other variables with smaller absolute sizes and standard deviations Advantages of non hierarchical clustering 1. More reliable & split sample will be working more similar to hierarchical clustering Both hierarchical & non hierarchical clusters are to be used in sequence I. First hierarchical clustering: Identifies the number of clusters and any outliers to obtain the cluster centers II. Non hierarchical clustering: outliers are removed and number of clusters are obtained from the hierarchical approach

Multidimensional scaling
It deals with the judgments of respondents about the degree of similarities of pairs of stimuli (objects). The similarities may be defined as the degree of resemblance between pairs. It is used to measure human perceptions & preferences towards some stimuli (objects) like products, organizations, places, events, brands, etc. MDS is ratio-scaled ,if the input & output data for MDS is metric. If the input & output data for MDS is non-metric ,MDS is ordinal-scaled. It involves three basic steps Identification of the dimensions of the perceptual space on which customers perceive objects. Input the respondent's preference of the object into MDS package. Positioning of the objects on the perceptual space. The output of MDS is a perceptual map which consists of locations of the objects in it. Applications of MDS Number & nature of dimensions consumers use to perceive brands in the market place Positioning of current brands on these dimensions Positioning of consumers ideal brand on these dimensions Assumptions & Limitations of MDS People perceive a set of objects as being more or less similar to one another on a number of dimensions assumed to be uncorrelated with one another. Because of computational & interpretational complexities MDS is not a popular technique in marketing research.

124

Business Research Methods An overview

125

Report Writing
GUIDELINES ON HOW TO WRITE A FORMAL REPORT Scientists and engineers routinely have the need to express themselves clearly, concisely and persuasively in applying for grants, publishing papers, reporting to their supervisors, communicating with their colleagues, etc.; in experimental science, formal reports are the primary means by which experimentalists communicate the results of their work to the scientific community. In the Formal Lab Reports, which are meant to be written in the style of a scientific journal article, you will have an opportunity to begin developing the skills required. Whereas the complete record of an experiment is kept in the lab book, the formal report is a summary of the experiment. Check out the Physics Library for examples - the American Journal of Physics is particularly useful and easily available. If you have doubts about your ability to write in clear English, you should check out the Writing Labs available in most of the Colleges. In a formal report the single most important thing is to write from the point of view of the reader. In the case of the scientific formal report you can safely assume that scientific journal readers are somewhat knowledgeable about science; perhaps for your own report you can assume that your readers will be first year physics students who have not done the particular experiment you are reporting. The reader of your report should understand from your report what you have done, why you have done it, and what you have concluded. The reader is not interested in going through the details of how you multiplied, divided, etc., and is unlikely to have any interest in long tables of numbers. You will have to use your judgment to determine what to include and what to exclude. For example, most common measuring apparatuses do not have to be described. However, any ingenious or novel method or tool in your experiment, should be explained in sufficient detail that the reader can understand what you did. In general the scientific paper is a very condensed, compact and brief presentation of scientific results. The report should be no more than 800 words, (3 pages of double spaced type) and 2 pages of graphs and/or diagrams. We will deduct marks if the report is longer than this! Please include your name, section and group, as well as your demonstrator's name. We recommend the following general structure:

125

Business Research Methods An overview

126

1.) Title: This should be short, but precise, and convey the point of the report. It could be either a statement or a question. For example, a title like "Voltage-current relationship of a transistor" is good, as is "Does the transistor obey Ohm's Law?". But simply "The transistor" is too vague and is not a good title. 2.) Abstract: The abstract summarizes, in a couple of sentences, the content of the report. It provides a brief (5-10 lines) outline of what the report is about; it should include a statement of what it is you measured and its value (Warning! -- students often make abstracts too long -- note that an abstract is not an introduction.) 3.) Introduction: The role of this section is to state why the work reported is useful, where it fits in the bigger picture of the field (or of science in general), and to discuss briefly the theoretical hypotheses which are to be tested (e.g. for the Absolute Zero experiment, state the meaning of absolute zero and how it is to be measured, mention the equation PV = nRT and discuss its verification, and under what circumstances you expect it to be valid). 4.) Experimental Method: Describe the apparatus and procedure used in the experiment. Remember that a picture (or simple diagram) is often worth a thousand words! Enough details should be provided for the reader to have a clear idea of what was done. But be careful to not swamp the reader with insignificant or useless facts. 5.) Results and Discussion: In this section, you present and interpret the data you have obtained. If at all possible, avoid tables of data. Graphs are usually a much clearer way to present data (make sure axes are labeled, and error bars are shown!). Please make sure the graphs and diagrams have concise figure captions explaining what they are about! Do not show the details of error calculations. The derivation of any formulae you use is not required, but should be referenced. Explain how your data corroborates (or does not corroborate) the hypotheses being tested, and compare, where possible, with other work. Also, estimate the magnitude of systematic errors which you feel might influence your results (e.g. In the Absolute Zero experiment, how big is the temperature correction? Does this alter your results significantly?). 6.) Conclusion: In a few lines, sum up the results of your experiment. Do your data agree (within experimental error) with theory? If not, can you explain why? Remember that the conclusion is a summary; do not say anything in the conclusion which you have not already discussed more fully earlier in the text.

126

Business Research Methods An overview

127

7.) References: In this section of the report list all of the documents that you refer to in your report. We recommend numbering the references sequentially in the text, in their order of appearance, and listing them in the same order in the references section. One possible reference format is used by the Canadian Journal of Physics. That is: for JOURNALS: Author(s), Journal Title, Volume (year) page number for BOOKS: Author(s), Book Title, Publisher, city of publication, year of publication, page (or range of relevant pages). Examples of reference entries: 1.) F. Azenberg-Selove, Nucl. Phys. A392 (1983) 1 2.) G. Zukav. The Dancing Wu Li Masters. Bantam, New York, 1979, p.23. Any report contains: descriptions on methodology, results obtained, and recommendations made. The basic orientation of a report depends on its audience. Before writing the report the researcher must know his or her audience; he/she may have to make assumptions about the composition, background and interests of the target readers. Two types of reports: Technical Report: suitable for a target audience of researchers, research managers or other people familiar with and interested in the technicalities such as research design, sampling methods, statistical details etc., Research oriented Report: suitable for: a more general audience, interested mainly in the research findings as it is non-technical in nature. The writing style is designed to facilitate easy and rapid reading and understanding of the research findings and recommendations

Prob mD fin ion le e it , Re a D sig se rch e n a dM t od y n e h olog

Dt aa An lysis a

Pre-report Writing Activities

In e re a ionof t rp t t Re a se rch find g in s

Re ort p Pre a t p ra ion

Report Writing Activities

O l ra Pre nt t ion se a t

RESEARCH FO W P LLO -U

Re d of t a ing he 127 y he Re ort b t p clie t n

Post Report Writing

Business Research Methods An overview

128

Components of Formal Reports

A formal report's manuscript format and impersonal tone convey an impression of professionalism. A formal report can be either short (fewer than 10 pages) or long (10 pages or more). It can be informational or analytical, direct or indirect. It may be written for internal or external audiences. What sets it apart from other reports is its polish. The parts included in a report depend on the type of report you are writing, the requirements of your audience, the organization you're working for, and the length of your report. The components fall into three categories, depending on where they are found in a report: prefatory parts, text of the report, and supplementary parts.

Prefatory Parts

128

Business Research Methods An overview

129

Many companies have standard covers for reports. If your company has no standard covers, look for a cover that is appropriate to the subject matter. The title page is a plain sheet of paper with only the title of the report on it. You don't really need one, but it adds a touch of formality. The title page includes four blocks of information: (1) title of the report; (2) name, title, and address of the person, group, or organization that authorized the report; (3) name, title, and address of the person, group, or organization that prepared the report; and (4) date on which the report was submitted. The letter or memo of authorization is a document requesting that a report be prepared. The letter or memo of acceptance acknowledges the assignment to conduct the study and to prepare the report. The letter or memo of transmittal conveys your report to your audience. The table of contents indicates in outline form the coverage, sequence, and relative importance of the information in the report. Some reports refer to all visual aids as illustrations or exhibits. In some, tables are labeled separately from figures. Regardless of the system used to label visual aids, the list of illustrations gives their titles and page numbers. A brief overview of a report's most important points is called a synopsis (or abstract in technical or academic reports). An executive summary is a fully developed "mini" version of the report.

Text of the Report

Intrduction Body Closing

Although reports may contain a variety of components, the heart of a report is always composed of three main parts: an introduction, body, and conclusion.

129

Business Research Methods An overview

130

A good introduction helps the reader follow and comprehend the information that follows. It invites the audience to continue reading by telling them what the report is about, why the audience should be concerned, and how the report is organized. This section contains the information that supports your conclusions and recommendations as well as your analysis, logic, and interpretation of the information. Most inexperienced writers have a tendency to include too much data in their reports or place too much data in paragraph format instead of using tables and charts. Such treatment increases the chance of boring or losing the audience. If you find yourself with too much information, include only the essential supporting data in the body, use visuals, and place any additional information in an appendix. In very general terms, you want the closing of your report to summarize the main idea of your report, highlight your conclusions or recommendations, and list any courses of action you expect the reader to take.

Supplementary parts

Supplementary parts follow the text of the report and provide information for readers who seek more detailed discussion. Supplements are more common in long reports than in short ones, and typically include the appendixes, bibliography, and index. An appendix contains materials related to the report but not included in the text because they're too lengthy or bulky or because they lack direct relevance. Frequently included in appendixes are sample questionnaires and cover letters, sample forms, computer printouts, statistical formulas, financial statements and spreadsheets, copies of important documents, and complex illustrations; a glossary may be put in an appendix or may stand as a separate supplementary part. The best place to include visual aids is in the text body nearest the point of discussion. But if any visuals are too large to fit on one page or are only indirectly relevant to your report, they too may be put in an appendix. You have an ethical and a legal obligation to give other people credit for their work. A bibliography is a list of secondary sources consulted when preparing the report. An index is an alphabetical list of names, places, and subjects mentioned in the report, along with the pages on which they occur (see the index for this book). An index is rarely included in unpublished reports. Proofreading Reports and Proposals

130

Business Research Methods An overview

131

Once you have assembled all the various components of your report or proposal, revised the entire documents content for clarity and conciseness, and designed the document to please readers, you have essentially produced your document in its final form. Now you need to review the entire document thoroughly one last time, looking for inconsistencies, errors, and missing components. Proofreading your report is pretty much the same as proofreading any business message check for typos, spelling errors, and mistakes in punctuation. However, reports often have elements that may not be included in other messages, so don't forget to proof your visual aids thoroughly. Make sure your text is laid out on the page in a clear, uncluttered fashion. Youll also want to make sure that nothing has been left out or overlooked, and that every word contributes to your report's purpose. Oral Reporting(presentation) Generally, an oral presentation supplements the written report. The client/audience should be given adequate time to read the report. If necessary, the researcher should assist the in understanding the report, implementing the findings, undertaking further research, and evaluating the research process in retrospect. Should be carefully prepared keeping the audience in mind. A good presentation does not mean a lengthy presentation. Carefully selected visual aids such as graphs, tables, charts, maps etc. help presentation. However, Too many visual aids, particularly statistical tables, could often be boring and may not serve any purpose. During oral presentation, people may seek clarification. The speaker must be patient and should not show signs of anger or frustration. He or she should be natural, establish eye contact with the audience, and interact with them. Body language and descriptive gestures are also quite useful.

131