Putting Statistics into Practice Strategies for effective management J Eric Davies & Claire Creaser IFLA Satellite 14/8/02 1 Options for measuring & managing J Eric Davies IFLA Satellite 14/8/02 2 Outline • • • • • Mission, vision, aims objectives Range of data types Applications of data Methods of acquiring data General principles IFLA Satellite 14/8/02 3 Measuring with meaning • Every organisation, no matter what its mission or scope needs three kinds of performance metrics • to measure its success in mobilizing its resources, • its staff’s effectiveness on the job, and • its progress in fulfilling its mission. – McKinsey Quarterly. 2001 -2 IFLA Satellite 14/8/02 4 Mission with meaning STRATEGIC FOCUS • WHAT • Organisation? NOW / LATER • Service? NOW / LATER • Direction? NOW / LATER • Mission / Vision • aims / objectives IFLA Satellite 14/8/02 5 Criteria AIMS/OBJECTIVES Specific Measurable Acceptable Realistic Time-bound {SMART} Consistent Testing {CUTE} IFLA Satellite 14/8/02 Unambiguous Empowering 6 How (im)possible is your mission? MISSION • Statement of purpose and functions – why service exists, what it does, who it serves VISION • Statement of desired future state – where service wants to be IFLA Satellite 14/8/02 7 How (im)possible is your mission? MISSION/VISION STATEMENTS ~ • • • • MEANING CREDIBILITY ACCEPTABILITY TESTABILITY IFLA Satellite 14/8/02 8 How (im)possible is your mission? MISSION/VISION STATEMENTS ~ Strathclyde University – Glasgow • A place of useful learning [1726] • The place of useful learning [2000] IFLA Satellite 14/8/02 9 IFLA Satellite 14/8/02 10 Here’s one I prepared earlier! MISSION/VISION STATEMENTS ~ • Public Library - The mission of the Library is to serve as a cathedral of human knowledge— an accessible database of knowledge that serves as the community's memory—and as an information and knowledge safety net, while providing materials, programs, and services to the people of the community. IFLA Satellite 14/8/02 11 Digging deep for evidence What Kind Of Evidence [Information]? • Statistics and Performance Indicators – [Quantitative + Qualitative ] • Social Measures – [Soft Indicators] ‘DISTANCE TRAVELLED’ IFLA Satellite 14/8/02 12 Who cares {or should do}? AUDIENCE: [stakeholders] • Funders • Managers/Staff • Users • Community • Vendors • Global IFLA Satellite 14/8/02 13 All kinds of measuring • Inputs • Outputs • Outcomes • Impacts IFLA Satellite 14/8/02 SERVICE DOMAIN s e r v i c e USER RESPONSES 14 Social dimensions Examples of ‘Soft’ Indicators:• • • • Attitudinal Personal Practical Key Work Skills DISTANCE TRAVELLED IFLA Satellite 14/8/02 15 Social dimensions • Personal development - individual selfconfidence, self awareness, creativity, new skills and abilities. • Social cohesion - Impact on group/community identity • Community involvement and empowerment • Health - people feeling better, happier etc. IFLA Satellite 14/8/02 16 How does the evidence add up? APPLICATIONS ~ SERVICES & PROJECTS • Policies • Strategies • Tactics • Processes and • Operations • Advocacy IFLA Satellite 14/8/02 17 What does the evidence answer? APPLICATIONS ~ SERVICES • How have we done? • How are we doing now? • How can we do better? • Where are we going? • How do we get there? • How are we making a difference? • How do we get the resources IFLA Satellite 14/8/02 18 What does the evidence answer? APPLICATIONS ~ PROJECTS • Did we achieve what we were seeking to achieve? • Did we do what we said we would do? • How did we do it? • What did we use? • What did we get out? • What worked and what didn’t work? • What could we do differently? • What can we apply continuously? • What difference did it make that we did it? • Who benefited? IFLA Satellite 14/8/02 19 Managing and measuring Framework for Performance Measurement:• integration • user satisfaction • effectiveness (delivery) • efficiency • economy Follett Report – academic libraries IFLA Satellite 14/8/02 20 Managing and measuring Three E’s – Economy in acquisition of resources – Efficiency in the use of resources – Effectiveness in the achievement of objectives UK Treasury [1980’s] FMI Sizer [1980’s] IFLA Satellite 14/8/02 21 Comparing and changing • BENCHMARKING – Motorola + D.E.C. + Xerox – To make changes that lead to quantum and continuous improvements in products, processes and services that result in total customer satisfaction and competitive advantage IFLA Satellite 14/8/02 22 Comparing and changing BENCHMARKING – • Evaluate the level of performance of various services within an institution • Overall level of institution performance • Compare against published standards • Compare performance over time • Compare with other institutions IFLA Satellite 14/8/02 23 Finding out Gathering Evidence • What do you need to know? • Where is the information? • Who has the information? • How will you get it? • How accurate is it / do you need it to be? • How will you interpret it? • How will you act on it? • How will you present it? IFLA Satellite 14/8/02 24 Gathering evidence Techniques/Tools/Options for Gathering Data:• MIS / Transaction Logs • Databases / Publications • Surveys : questionnaire,telephone, interview • Focus Groups / Graffiti boards • Observation / Diaries / Logs • Press~Media Coverage IFLA Satellite 14/8/02 25 Gathering evidence TOOLS:• What Outcomes, Dimensions, Performance to be measured? – reliable + valid – meaningful and precise IFLA Satellite 14/8/02 26 Gathering evidence Options for Gathering Data • … if the only tool you have is a hammer, everything starts looking like a nail. F.W. Huibregston - Partner: McKinsey’s IFLA Satellite 14/8/02 27 Changing times; changing evidence UPDATING EVIDENCE • Service Evolution – New/Discontinued services - methods technologies - clients • Diminishing Variance – Improvement - Gaming - Deception IFLA Satellite 14/8/02 28 Manager beware!! OVERDOING IT: If you know everything, you know nothing George Johnson: Fire in the Mind. [1996] … a world that never measures or counts is really beyond our control. The trouble is that we’re in danger of doing little else. David Boyle: RSA Lecture . [2001] IFLA Satellite 14/8/02 29 How much evidence? ... data is not information. Information is data endowed with relevance and purpose. A company must decide what information it needs to operate its affairs, otherwise it will drown in data Peter Drucker - Managing for the Future. IFLA Satellite 14/8/02 30 Making sense of measuring Sumsion’s Law of Statistical Dullness ~ In comparative statistics the great majority of results are inherently close to the average and consequently dull. {Sumsion} LIRN 2001 (79) p.3. IFLA Satellite 14/8/02 31 Evidence for yesterday Statistics, being essentially historical, can only provide information after the event. {Sumsion} Int. Encyclopedia of Lib. and Info. Sci.(1997)p.432 IFLA Satellite 14/8/02 32 Measuring and managing Information is a precondition for identifying choices, reducing uncertainty about their implications and facilitating their implementation. Center for Transnational Corporations:- CTC Reporter 14 Winter 1983 p.34 - IFLA Satellite 14/8/02 33 Managing and measuring; comparing and changing • LISU: We’ve got the measure of information! – A skilled team of experienced Managers, Statisticians and Administrators all adding value to statistical data and providing authoritative and reliable information to support managers in culture, information and related environments. IFLA Satellite 14/8/02 34 Library and Information Statistics Unit lisu@lboro.ac.uk IFLA Satellite 14/8/02 35 Mission possibilities • Does it have – Meaning? • Does it actually mean anything? – Credibility • Do you believe it can be achieved? – Acceptability • Will all the stakeholders (funders, staff, users) ‘buy-in’ to this mission? – Testability • How would you demonstrate you are achieving your mission? IFLA Satellite 14/8/02 36 Library and Information Statistics Unit lisu@lboro.ac.uk IFLA Satellite 14/8/02 37 Statistics for the faint hearted Claire Creaser CStat IFLA Satellite 14/8/02 38 Introduction to statistics • Basics – What are statistics – Useful techniques • Sampling – Surveys and sample sizes – Questionnaire design – Analysis • Benchmarking • Presentation of results IFLA Satellite 14/8/02 39 What are statistics? • Numbers with context – – – – 1,300 items issued last month The average price paid for a CD is £12.50 25% of staff time is spent re-shelving books Women borrow twice as many books on average as men – Serials cost three times as much as books – The average spend per user has increased less than general inflation over the last ten years IFLA Satellite 14/8/02 40 Where to start • What do you want to know? – Evidence of good management – Value for money – Advocacy • What data to collect? – What do you want to know? • Relevant • Useful • Current IFLA Satellite 14/8/02 41 Where do they come from? • Library management systems – Stock statistics, financial data, staff . . . . • Regular surveys – User opinions, condition of stock . . . . • Occasional surveys – Project evaluation IFLA Satellite 14/8/02 42 What do they look like? • Categorical – Gender; classmark; membership status • Ordinal – Stock condition; satisfaction ratings • Ratio or interval – Acquisitions; issues; expenditure IFLA Satellite 14/8/02 43 What can you do with them? • What do you want to know? • Descriptive statistics – Mean, range, distributions, proportions – graphical presentations • Inference from samples – Estimates, error levels • Advanced techniques – Correlation, regression, analysis of variance IFLA Satellite 14/8/02 44 Choosing the right technique • Keep it simple! • Categorical data – Proportions in each category – Comparisons • Ordinal data – Proportions in each category – Medians • Ratio data – Means IFLA Satellite 14/8/02 45 Sample surveys • Why sample? – Cost – Practicalities • Where to start? – Sampling frame – Sample design – Sample size IFLA Satellite 14/8/02 46 Types of sample • • • • • Simple random Systematic Stratification and clustering Quota samples Self-selected IFLA Satellite 14/8/02 47 How many? • Less than you think! • Depends on: – Level of detail – Desired margin of error – Expected response rate • Does not depend on population size – Unless small population • 400 will give accuracy of ± 5% • 1,000 for ± 3% • 2,500 for ± 2% IFLA Satellite 14/8/02 48 Questionnaire design • • • • Self-completion or interview? Clear, unambiguous questions Clear, easy to follow layout As short as possible – Number of questions – Number of pages • Tick boxes or short answers • Data entry issues IFLA Satellite 14/8/02 49 Sampling times • One period, or several? • Periodicity IFLA Satellite 14/8/02 50 Analysis • Lots of software available – Spreadsheets • Excel, Lotus – Specialist packages • SPSS, Minitab • Keep it simple! • If in doubt – bring in a statistician! IFLA Satellite 14/8/02 51 Making comparisons - benchmarking • Compare like with like! • Within your service – Performance of service points – Behaviour of user groups • Externally – National standards – Other libraries • Over time – Trends IFLA Satellite 14/8/02 52 Presentation of results • Appropriate to – The data type – The audience • Tables or graphs? • Key features: – – – – Clarity Integrity Descriptive Appropriateness IFLA Satellite 14/8/02 53 Graphical forms • Line graphs – For presenting trends over time • Pie charts – For presenting proportions of a whole • Bar or column charts – For comparing categories • Scatter graphs – For examining relationships IFLA Satellite 14/8/02 54 Issues per capita in UK public libraries 300 250 200 Adult nonfiction 150 100 Children's AV 50 99-00 98-99 97-98 96-97 95-96 94-95 0 93-94 26 28 30 31 33 34 36 37 37 38 38 Adult fiction 92-93 105 105 113 112 112 112 110 111 111 109 101 350 91-92 AV 90-91 333 326 325 314 303 289 275 261 246 234 221 Children's 89-90 89-90 90-91 91-92 92-93 93-94 94-95 95-96 96-97 97-98 98-99 99-00 Adult nonfiction 129 129 138 136 136 133 129 127 122 114 107 millions Adult fiction Source: LISU Annual Library Statistics IFLA Satellite 14/8/02 55 Breakdown of UK academic library expenditure Staff Information Of which: Books Serials Electronic ILL Binding Equipment Other £128m £92m £27.0m £48.6m £8.4m £3.5m £4.9m £20m £17m Equipment 8% Other 7% Serials 19% Electronic 3% Information provision 36% ILL 1% Staff 50% Books 10% Binding 2% Source: LISU Annual Library Statistics IFLA Satellite 14/8/02 56 Activity in public libraries Proportion of visitors who: Borrow/return books 74.2% Borrow return cassettes 7.0% Borrow/return CDs 4.9% Borrow/return videos 8.4% Read newspapers 13.8% Seek information 22.9% Use photocopier 6.0% Browse 29.9% Study 7.7% Visit event 2.9% Other 7.8% Borrow/return books Browse Seek information Read newspapers Borrow/return videos Study Borrow return cassettes Use photocopier Borrow/return CDs Visit event Other 0 20 40 60 80 % of visitors Source: CIPFA Public Library User Survey IFLA Satellite 14/8/02 57 Issues 630,788 472,514 424,045 413,133 474,982 297,387 328,578 225,841 267,913 98,050 70,349 113,049 0.7 0.6 0.5 Issues Acquisitions A1 12,242 B1 10,730 B2 7,184 B3 8,121 B4 6,075 C1 2,895 C2 4,155 C3 3,572 C4 4,940 D1 1,488 D2 1,435 D3 2,471 ……… Millions Public library book acquisitions and issues 0.4 0.3 0.2 0.1 0.0 0 2 4 6 8 Acquisitions 10 12 14 Thousands Source: UK public library authority data, unpublished IFLA Satellite 14/8/02 58 Library and Information Statistics Unit lisu@lboro.ac.uk IFLA Satellite 14/8/02 59 Managing scenarios, strategies and statistics in practice • • • • • • • What do you need to know? Where is the information? Who has the information? How will you get it? How will you interpret it? How will you act on it? How will you present it? IFLA Satellite 14/8/02 60 Library and Information Statistics Unit lisu@lboro.ac.uk IFLA Satellite 14/8/02 61 ISI Glossary Access from http://europa.eu.int/comm/eurostat/research /index.htm IFLA Satellite 14/8/02 62