Analysing-quantitative-data.ppt

Analysing dataQualitative v quantitative dataQualitative data*uses open ended questions to get in depth answers*avoids closed yes/no questions*idea is that people should share their thoughtsQuantitative data*gather numerical information*sample sizes larger*usually questionnaire research where results can be summarised numbers percentages averages*involves yes/no,Likert scale,etcBasic comments on data*choose the easiest way to analyse your data*keep it simple*choose the data presentation method that best communicates the information1.Preparing the data for analysis *Data types *data can be put into categories e.g.hatchback cars saloon cars estate cars BUT make sure the categories do not overlap *data can be quantifiable e.g.you give a position on an attitude scale a number*Data collection*if you issue a questionnaire on line and it is returned on line analysis of data can be easierBUT many questionnaires are returned manually,so:-*you have to prepare data *you have to code data*Coding data*with numbers,use the numbers when coding e.g.employee salary 32,000 code it 32,000 analyse by group 30-40,000*with data from open ended questions,coding can be done after you have collected the data as you will be unsure of the likely responses.Thus:-*wait for 20-30 responses *look at responses and develop broad groupings *divide the broad groupings into sub groups *allocate codes *code further responses e.g.What are the key factors that might improve your degree?Answers less hours more free time more lectures more books fewer assignments better lecturersCould categorise these as Academic Non-academicOR Factors that can be changed quickly Factors that can not be changed quickly 2.Presenting your dataYou need to remember what your research objectives areBUT your data may also show results you had not anticipatedWhat do you look for to get started?How do you present your data?Highest and lowest valuesshown as a bar chartStudent attendance at seminarsTrendsused for data collected over timecan be shown in a graphProportion of population in a categorycan be shown in a pie chart Students attending each seminarYou may also want to make comparisonsHighest and lowest comparisonsYou can use a bar chartStudent attendance per group each weekYou can also compare trendsStudent attendance by group each week3.Describing data using statistics How do you describe your data using statistics?Here you will be discussing such as:-A.Central tendency i.e.most common,middling or average value *mode:the value that occurs most frequently e.g.the most common(modal)colour for new cars last year was red*median the middle value of a distribution*mean or average BUT this might be skewed by a few extreme valuesExampleFamily Income()1 25,000 Mean 55,000(average)2 30,0003 30,0004 30,000 Mode 30,000(most5 30,000 frequent value)6 40,0007 40,0008 45,000 Median 35,000(middle 9 50,000 value)10 230,000Which calculation makes most sense?B.Dispersion around the central tendencyThis is the distribution of the data*a quick calculation is to look at the difference between the highest and lowestBUT this may not have much meaning*instead you can use quartiles *lower quartile-value below which a quarter of your data falls *upper quartile-value above which a quarter of your data falls*you may want to describe how data is spread around the mean.If the data values are all close to the mean,then the mean is more typical than if they vary widely.To describe the extent of spread of quantifiable data you can use the standard deviation 4.Looking at the relationships in the dataThe probability is of your test results occurring by chance These are tests of validityYou will not be able to do this if your sample is small One test of whether two variables are associated isChi squared testA predictability of 0.05 means there is only a 5%chance of your data occurring by chance,and thus you can be 95%certain that the relationship between your two variables could not have occurred by chance.The results are then statistically significant.(see Sanders et al)There are other calculations such asregression analysisThere is no requirement to do the above calculations -it depends on the size of your sampleBut be aware of the limitations of your data if you do not do the calculations Main problem is students describe the data onlyThe project write upHow do you produce an acceptable written project?1.Getting started*last minute?*it can take a day to write 2000 words*some people write till they drop*put aside 3 or 4 hours a day*plan*write when you are fresh*regular place to write BUT avoid disruptions distractions*music*set targets e.g.number of words*set a word count on your pcProject structureTitle pageList of contentsAbstract *a short summary *will be read most *abstract should contain the following:-what were the research questions?why were they important?what was the research methodology?what did you find out?what were your conclusions?Chapter 1 Introduction *should give background context *not a long chapter Chapter 2 The issue/problem to be investigated *what it is *how it arose *why it is important Chapter should conclude with clear project aims and objectives.Chapter 3 Literature review *wider context *ideas and research of others *conclusion *is a key chapter *needs to go beyond description into analysis and evaluationChapter 4 Research methods *outline and justify your research methods *this chapter should explain how you carried out the research why you carried out the research *this chapter can include why you chose the sector/organisation what ethical issues were raised how you addressed these how many participants there were how they were selected*this chapter can include collection methods questionnaire/interview schedule how you tested it how the results were analysed when the research was carried outChapter 5 Data analysis and evaluation *present your findings *may include tables,graphs,etc *use quotes from respondents if this gets your point across *begin to evaluate and make connections Chapter 6 Conclusions and recommendations *key chapter *do not just repeat findings *you need to ensure you address your aims and objectives Final PageWhat I learnedReferencesAppendices *not part of word countSome thoughts on a good project*choose a title the best use few words*help the reader sub headings preview and summarise chapters use tables and graphs refer to these put them at right place in text *help the reader write simple understandable sentences avoid jargon do not use large numbers of quotes from the literature check grammar and spelling*assume the reader knows nothing*dont wait until the end to start writing up you can write background to industry/organisation issue/problem*write the literature review clarifies your thoughts *develop a good writing style write in the past tense do not use I use the researcher*be aware of gender issues*revise your work continually produce draftsFinal thoughts *one report or two?you may be producing *one academic report longer contains organisation background has literature review *one for managers shorter no background or literature concentrates on findings and recommendationsUnderstand the marking criteriaPROJECT PROPOSALYou will need to write and submit this1000 wordsCriteria for evaluating the research proposal *is the topic appropriate to the degree?*can it be researched in time?*are the aims and objectives clear?*are the aims and objectives researchable?*is there a clear link to theory?*are the research methods appropriate?*are the findings likely to be of interest to an organisation or organisations?