Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.1 Chapter 16: Analysing Survey Data
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.2 Contents Survey data analysis and types of research Spreadsheet analysis Statistical package for the social sciences (SPSS) Preparation SPSS procedures The analysis process
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.3 Research typeAnalytical procedures Descriptive Frequencies, means Explanatory Crosstabulation, comparison of means, regression Evaluative Frequencies – compared with targets or benchmarks Crosstabulations – comparing user/customer-groups Means – compared with some benchmark or target Figure 16.1 Survey data analysis and types of research
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.4 Explanatory research and causality Necessary conditions: –Associations between variables (A changes with B) –Time priority (B happens after A) –Non-spurious relationships (relationships ‘make sense’) –Rationale/theory (there should be an explanation)
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.5 Example using data from Campus Life questionnaire (Figure 10.21) FREQUENCY procedure in Microsoft Excel used to produce: –frequency counts of coded variables –averages for numerical variables (age, spend) Figure 16.2 Spreadsheet analysis
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.6 Statistical package for the social sciences (SPSS) Software package produced by SPSS Inc., owned by IBM Can be used to analyse questionnaire-based and other data organised as cases with specified variables SPSS is effective and one of the most popular packages. Its use in this book does not imply endorsement as ‘the best’ package
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.7 SPSS procedures covered Figure 16.4 Survey analysis – overview 3. Descriptives 4. Multiple response 5. Recode 7. Weighting 6. Means 8. Crosstabs 9. Statistics - see Chapter Graphics 1. Preparation 2. Frequencies
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.8 Preparation: cases and variables: from Figure Variables qnostatuscafebarmusicsporttravelcheapEtc. CasesCases
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide 16.9 Information required for each variable in the questionnaire Name Type – numeric, string (letters) or date Width – max. no. of characters Decimal places Label – longer version of name Values for coded variables Missing – blanks, no answer, etc. Columns – no. of columns in Data view screen (see below) Alignment – left, right, centre (in Data View) Measure/data type – nominal, ordinal, scale
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Variable names Up to 8 characters (no spaces), beginning with a letter Not allowed: ALL AND BY EQ GT LE LT NE NOT OR TO WITH Can be: –Short version of item description (as used here), or –var01, var02, var03, etc. or –Q1a, Q1b, Q2, Q3, etc.
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Types of measure Nominal: Described in words – e.g. male/female Ordinal: Ranked: 1, 2, 3… means 1st, 2nd, 3rd… Scale: Fully numeric
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Variable View Information on variables is entered in the SPSS ‘Variable View’ screen
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Variable View screen
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Data View Data entered directly on the Data View screen, or Can be imported from a spreadsheet file
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Data View screen
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Note to teachers It is not envisaged that SPSS detailed procedures would be the subject of a PowerPoint presentation: students would benefit most from following the procedures in practical sessions A copy of the Campus Life data files is available on the book website However, teachers may wish to discuss the nature/ purpose of the various procedures Slides are therefore included with the outputs from the procedures
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Descriptives: N, Minimum, Maximum, Mean & Standard Deviation for each variable
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide NMin.Max.Mean Std. Deviation Student status Campus cafe/bar in last 4 wks Live campus music in last 4 wks Sport facilities in last 4 wks Travel service in last 4 wks Free/cheap (rank) Daytime events (rank) Not available elsewhere (rank) Socialising (rank) Quality of presentation (rank) Entertainment expenditure/month Relaxation opportunities – importance Etc. Figure Descriptives: output: first few variables
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Frequencies Simple counts/percentages of variables Nominal/ordinal: straightforward Numeric may need to be grouped – see Recode Frequencies form the basis for a statistical summary/appendix – see Figure 16.6
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Student statusFrequencyPercent Valid percent Cumulative percent ValidF/T student/no paid work F/T student/paid work P/T student – F/T job P/T student/Other Total Frequencies for all variables: see Appendix 16.1 Figure Frequencies: output
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Multiple response Two types of ‘Multiple Response’ Dichotomy: Q.2: Use of services: 4 ‘yes/no’ variables –Best combined into one table Category: Q.6: Suggestions: up to three responses per respondent = 3 variables –Best combined into one table
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Dichotomy label NameCountPct of Responses Pct of Cases Campus cafe/bar in last 4 wkscafebar Live campus music in last 4 wksmusic Sport facilities in last 4 wkssport Travel service in last 4 wkstravel Total responses Category label CodeCount Pct of Responses Pct of Cases Programme content Timing Facilities Costs Organisation Total responses Figure Multiple response output
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Recode Grouping/Re-grouping variable categories, especially: –presentational: numerical variables –theoretical e.g. 5 categories of tourism or just two: leisure vs non-leisure? –Comparison – with other research –statistical reasons – see Chapter 17 Examples: – Uncoded, ‘spend’ has 9 different answers (see Appendix 16.1): recode into 4 groups –Student status has 2 F/T and 2 P/T categories: recode into F/T and P/T
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Spend recodedFrequencyPercentValid percent Cumulative percent £ 0 – £ 51 – £ 101 – £ Total Status recodedFrequencyPercentValid percent Cumulative percent Full-time student746.7 Part-time student Total Figure Recode: output
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Measures of central tendency: Mean, median, mode Mean = average Median = middle value when all cases ranked in order Mode = most popular value Only valid with scale and ordinal variables Options: –Add to ‘Frequencies’ procedure –Use ‘Means’
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Mean, median, mode: using ‘frequencies’ procedure Additional output from ‘Frequencies’ Relaxation opportunities – importance Social interaction – importance Mental stimulation – importance NValid 15 Missing 000 Mean Median Mode 231
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Means procedure Student statusMeanNStd. Deviation* F/T student/no paid work F/T student/paid work P/T student – F/T job P/T student/other Total Mean expenditure by student status
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Crosstabulation Table showing relationships between two or more variables Table can include one or more of the following: – Counts –Row % –Column % –Total % –Statistical tests – see Chapter 17 Procedure: ‘Crosstabs’
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Crosstabs Student status by attended live campus music: counts only Live campus music in last 4 wks Total NoYes Student status F/T student/no paid work 112 F/T student/paid work 325 P/T student - F/T job 246 P/T student/Other 112 Total 7815 Student status by attended live campus music: row percentages Live campus music in last 4 wks Total NoYes Student status F/T student/no paid work 50.0% 100.0% F/T student/paid work 60.0%40.0%100.0% P/T student - F/T job 33.3%66.7%100.0% P/T student/Other 50.0% 100.0% Total 46.7%53.3%100.0%
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Crosstabs (Continued): three variables Gender Live campus music/4 wks Total NoYes MaleStudent status F/T student/no paid wk 112 P/T student - F/T job 235 P/T student/Other 011 Total 358 FemaleStudent status F/T student/paid work 325 P/T student - F/T job 011 P/T student/Other 101 Total 43 7
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Weighting Weighting discussed in Chapter 13 ‘Weight cases’ procedure e.g. if Masters students under-sampled: –suppose masters students need to be given a weight of 1.3 –create new variable wt –for Masters students wt = 1.3; all others: wt = 1 –In ‘Weight cases’: weight by wt
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Graphics Types: –bar graph –stacked bar graph –pie chart –line graph –scatter plot Different graph types suited to different data types
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Data type NominalOrdinalScale Data characteristicsQualitative categories RanksNumerical Example questions in Figure , 2, 6,73, 54, 8 Mean/average possibleNoYes Types of graphic Bar graphYes Yes* Pie chartYes Yes* Line graphNo Yes Scatter plotNo Yes * Grouped – Figure Data types and graphics
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Figure Bar chart
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Stacked bar chart
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Pie chart
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Line graph
Tony Veal, Research Methods in Leisure and Tourism, 4 th Edition © Pearson Education Limited 2011 Slide Scatterplot