1、Quantitative Business Analysispage:1Objektives of the coursepage:21.To gain an understanding of the importance of qba for business,economics and finance2.To develop spread-sheet skills applicable to the academic and business world3.To be able to present data in a meaningful way and to interpret char
2、ts and displays of data4.To know the basic tools of descriptive and inferentialstatisticsIntroduction:Defining the Role of Statistics in BusinessChapter 1page:31.Analyzing Business Data-Introduction1.1.We are Zalando/Amazon,what statistical data do we use?1.2.Compile an excel-file with your personal
3、 data on-NameCountry of originHeight Shoe size Gender Favorite sports Favorite foodFavorite internet-siteFavorite Brand2page:41.Analyzing Business Data-Introduction1.1.Business Statistics What is business statistics?Collecting,organizing,quantitative data.analyzingandinterpretingqualitativeor Why do
4、 we need statistics?Statistics help make better decision by understandingvariation and by uncovering patterns and relationships inFour Steps:the source ofbusiness data.1.2.Present and describe business data and information properlyDraw conclusions about large groups of individuals or items,using inf
5、ormation collected from subsets of the individuals or itemsMake reliable forecasts about a business activityImprove business processes3.4.2page:51.Analyzing Business Data-Introduction1.2.Quantitative AnalysisInputProcessOutputInformation Quantitative analysis is a scientific approach to managerial d
6、ecisionmaking whereby raw data are processed and manipulated resulting inmeaningful information.Quantitative analysis provides data-driven analytical services for a range of business challenges,specializing in statistical models for site selection decisions.Examples:When to order additional new mate
7、rial?What is the safety stocklevel?3page:6MeaningfulQuantitativeAnalysisRaw Data1.Analyzing Business Data-Introduction1.3.Descriptive and Inferential Statistics(I)We differentiate between descriptive and inferential statistics Descriptive statistics include techniques that are useddescribe numerical
8、 data for easier interpretation(e.g.into summarize andtables or graphs).Determination of parameters e.g.average,median or variances Inferential statistics include techniques by which decisions about astatistical population or process are made only on a sample having beenobserved with hypothese testi
9、ng and estimations=the resulting decisions are made under conditions of uncertainty.Therefore the use of probability is requiredpage:71.Analyzing Business Data-Introduction1.3.Descriptive and Inferential Statistics(II)page:8InferentialDrawing conclusions and/or making decisions concerning a populati
10、on based only on sample dataDescriptiveCollecting,summarizing and describing dataStatisticsData and Data SetsData are the facts and figures collected,analyzed,and summarized for presentation and interpretation.All the data collected in a particular study are referredto as the data set for the study.
11、page:9Elements,Variables,and ObservationsElements are the entities on which data are collected.A variable is a characteristic of interest for the elements.The set of measurements obtained for a particularelement is called an observation.A data set with n elements contains n observations.The total nu
12、mber of data values in a complete data set is the number of elements multiplied by the number of variables.page:10Data,Data Sets,Elements,Variables,and ObservationsVariablesElementNamesCompanyData Setpage:11Dataram EnergySouth Keystone LandCarePsychemedicsNQ73.100.86N74.001.67N365.700.86NQ111.400.33
13、N17.600.13StockAnnualEarn/ExchangeSales($M)Share($)Scales of MeasurementScales of measurementNominalinclude:IntervalOrdinalRatioThe scale determines the amount of informationcontained in the data.The scale indicates the data summarization andstatistical analyses that are most appropriate.page:12Scal
14、es of MeasurementNominalData are labels or namesattribute of the element.used to identify anA nonnumeric label or numeric code may be used.page:13Scales of MeasurementNominalExample:Students of a university are classified by theschool in which they are enrolled using anonnumeric label such as Busine
15、ss,Humanities,Education,and so on.Alternatively,a numeric code could be used forthe school variable(e.g.1 denotes Business,2 denotes Humanities,3 denotes Education,so on).andpage:14Scales of MeasurementOrdinalThe data have the properties of nominal data andthe order or rank of the data is meaningful
16、.A nonnumeric label or numeric code may be used.page:15Scales of MeasurementOrdinalExample:Students of a university are classified by theirclass standing using a nonnumeric label such asFreshman,Sophomore,Junior,or Senior.Alternatively,a numeric code could be used for the class standing variable(e.g
17、.1 denotesFreshman,2 denotes Sophomore,and so on).page:16Scales of MeasurementIntervalThe data have the properties of ordinal data,andthe interval between observationsterms of a fixed unit of measure.is expressed inInterval data are always numeric.page:17Scales of MeasurementIntervalExample:Melissa
18、has an SAT score of 1885,while Kevinhas an SAT score of 1780.Kevin.Melissa scored 105points more thanpage:18Scales of MeasurementRatioThe data have all the properties of interval dataand the ratio of two values is meaningful.Variables such as distance,height,weight,and timeuse the ratio scale.This s
19、cale must contain a zero value thatthat nothing exists for the variable at theindicateszero point.page:19Scales of MeasurementRatioExample:Melissas college record shows 36 credit hoursearned,while Kevins record shows 72 credithourshoursearned.earnedKevin has twice as many creditas Melissa.page:20Cat
20、egorical and Quantitative DataData can be further classified as beingor quantitative.categoricalThe statistical analysis that is appropriate dependson whether the data for the variable are categoricalor quantitative.In general,there are more alternatives foranalysis when the data are quantitative.st
21、atisticalpage:21Categorical DataLabels or names used to identify an attributeeach elementofOften referred to as qualitative dataUse either the nominal or ordinal scalemeasurementofCan be either numeric or nonnumericAppropriate statistical analyses are rather limitedpage:22Quantitative DataQuantitati
22、ve data indicate how many or howmuch:discrete,if measuring how manycontinuous,if measuring how muchQuantitative data are always numeric.Ordinary arithmetic operations are meaningfulquantitative data.forpage:23Scales of MeasurementDataQuantitativeCategoricalNumericNumericNon-numericIntervalRatioNomin
23、alOrdinalNominalOrdinalpage:24Cross-Sectional DataCross-sectional data are collected at the same orapproximately the same point in time.Example:data detailing the number of buildingpermits issued in November 2010 in each of thecounties of Ohiopage:25Time Series DataTime series data are collected ove
24、r several timeperiods.Example:data detailing the number of buildingpermits issued in Lucas County,Ohio in eachthe last 36 monthsofGraphs of time series help analysts understandwhat happened in the past,identify any trends over time,andproject future levels for the time seriespage:26Time Series Data0
25、7.08.group 1+2Series DataGraph of TimeU.S.Average Price Per GallonFor Conventional Regular GasolineSource:Energy Information Administration,U.S.Department of Energy,May 2009.page:27Statistics in Business:ExamplesAdvertisingEffective?Which commercial?Which markets?Quality controlDefect rate?Cost?Are
26、improvements working?FinanceRisk-How high?How to control?At what cost?AccountingAudit to check financial statements.Is error material?What sample-size do I need if basesd on system audit?OtherEconomic forecasting,background info,measuring and controllingproductivity(human and machine),page:28Modern
27、Finance and Accounting rely on statisticsPortfolio-ManagementStandarddeviation is main meassure of riskCorrelation-coefficients are meassured to reduce riskCorporate ValuationVaRRegression-analysis is used to determin Beta,a meassure of market riskValue at Risk(VaR)modells are used to manage and rep
28、ort the range ofpossible outcomes for financial instruments and are based on statisticalanlyses of historic price movementsOption ValuationOption Pricing modells typically include meassures of variability.Scoring-ModellsStatistical modells to predict whether or not a borrower will defaultpage:29Coll
29、ecting DataProject:Bike usage predictionChapter 2page:30Project:Bike usage predictionYou are provided hourly rental data spanning two years.Thetraining set(train data)is comprised of the first 15 days of each month,while the test set(test data)coverst the days 16 to 19 of each month.You are the busi
30、ness analyst of the company and managementhas given you the task to wirite a report covering the followingpoints:1)2)3)Describe statistically the data available to managementShow useful associations of the indivdual variables Develop a model that allows management to predict the required bikes depen
31、ding on some or all of the variables availabelUse the test data set to make predictions of bike usage based on the data availableAssume you are provided with the actaul usage data of your test data.How would you evaluate the accuracy of your prediction model?4)5)page:31Project:Bike usage predictionp
32、age:322.How to Obtain Statistical Data2.1.Ways of Obtaining DataStatistics is a tool for converting data into informationQuestions:Where then does data come from?How is it gathered?How do we ensure its accurate?Is the data reliable?Is it representative of the population from which it was drawn?page:
33、33SurveyDirectObservationThere are two ways of obtaining dataTypes of Research,Methods and TechniquesTypes of research(determined by its purpose)Methods and techniques of researchAnalysis of primary data(collected during researchproject)or secondary data(taken from print orelectronic media)Data coll
34、ected from a sample(sample survey,sampling inspection)or from the whole population(census,inventory)qualitative methodsexploratoryto get an insightidea or gaindescriptive to gather basicinformationexplanatoryexperience surveys(expert interviews)case studies(investigation of single cases)to explain r
35、elationshipspilot studies depth interviews)(unstructured,focus group,in-causal sample surveys(based on non-random sampling)quantitative methods to determinecausationobservations(measurements)experiments(variation of controlled parameters)sample surveys(based on random sampling),structured interviews
36、,questionnairespage:34evaluation research to assess results of activities/programs2.Analyzing Business Data-Introduction2.2.Direct Observation Statistical data can be ontained by direct observation e.g.when samplesof output are systematically assessed.Another form of direct observations is a statist
37、ical experiment,in whichthere is controlstudied.over some or all factors that may influence the variable beingpage:352.Analyzing Business Data-Introduction2.3.Survey If statistical data cannot be collected directly,a statistical survey can bemade.This is the process of collecting data by asking indi
38、viduals to provide the data.A survey solicits information from people;e.g.marketing surveys.The response rate(i.e.the proportion of all people selectedthe survey)is a key survey parameter.who complete Surveys may be administered in a varietyPersonal Interview,Telephone Interview,of ways,e.g.Self Adm
39、inistered Questionnaire,Internetandpage:36(arbitrary/aimlessly)Quantitative Methods:Sample SurveysSample survey methods(indirect methods of obtaining data):Questioning part of a population(a sample)-as opposed to a census(that investigates the whole population).Different sampling methods:Non-random
40、sampling Random sampling purposive or judgemental simple random samplingsampling systematic r.s.convenience sampling stratified r.s.cluster r.s.Random sampling means a way of selecting the sample such that anyindividual has equal chance to be selected.Only from randomly selected samples,the obtained
41、 data may be generalized to the whole population!aimlessly is not the same as randomly!page:37Random Sampling Techniques Examples Simple random sampling Lotto game(“6 out of 49“)and other lotteries choosing 100 students of FH Furtwangen(using randomnumbers/students list)Systematic random sampling Ch
42、oosing a random no.N between 1 and 100 and all Students list with No.N+k*100 where k=0,1,2,Stratified random samplingon the Choosing samples of male and of female students the sample sizes ofwhich are proportionate to the sizes of the male and female student population Choosing samples of male and f
43、emale German and Non-German students the sample sizes of which are proportionate to the sizes of the male and female German and Non-German student populationCluster random sampling Choosing a number n of students dormitories(ok only,if variable in question is not likely to depend on living inapage:3
44、838dormitory or not)Objectivity Validity Reliability Representativeness of DataObjectivity/Precision of data no subjective influencesQuestionnaire:no suggestive questioning/no influence by othersObservation:no systematic error of measuring instrumentValidity of data data measure what they are meant
45、toQuestionnaire:precise wording of questions/precise wording of answersObservation:accuracy of measuring instrumentReliability of data data measurement is reproducible under equal conditionsQuestionnaire:equal questions get same answers at different timesObservation:consistency of measuring instrume
46、ntRepresentativeness of data data collected from a sample that represents the entire target populationQuestionnaire:random samplingObservation:random samplingpage:39Statistical Presentation and Graphical DisplaysChapter 3page:40Principals of Graphical ExcellenceGraphical excellence requirespresentat
47、ion of data thata well-arranged12.10.providessubstancestatisticsdesigncommunicates complexprecisionideas with clarity andgives the largest number of ideas in the most efficient manneralmost always involves several dimensionsrequires telling the truth about the datapage:41S a vi n g s100 60 20 page:4
48、245120 40 35 3080 25 20 1540 10 50 0S t o c k sB o n d sS a vi n g sC D C D B o n d sS t o c k s01020304050 pie chartspareto diagrambar chartstabulating dataSummary Tablegraphing datacategorical dataOrganizing Categorical DataVariables are Categorical.(CD=certificate of deposit)page:43Investment Cat
49、egoryAmountPercentageStocks Bonds CD SavingsTotal(in thousands$)46.53215.51642.2729.0914.0914.55110100Example:Summary Table for an investors portfolioOrganizing Categorical Datapage:44Investors PortfolioSavings CD Bonds Stocks01020304050Amount in K$Bar Chart(for an investors portfolio)Graphing Categ
50、orical Data14%29%page:45BondsPercentages are rounded to the nearest percent.CDStocks42%Savings15%Amount Invested in K$Pie Chart(for an investors portfolio)Graphing Categorical Datagraph showsthe order ofhorizontal2040Group 1+2:18.10.Pareto DiagramAxis for barchart shows%invested in each category sta
