Management Math for Libraries An Infopeople Workshop Jeanne Goodrich [email protected] Winter 2004-05 AGENDA • Numbers Issues • Percentages and Trends • Descriptive Statistics • Making Management Decisions • Your.
Download ReportTranscript Management Math for Libraries An Infopeople Workshop Jeanne Goodrich [email protected] Winter 2004-05 AGENDA • Numbers Issues • Percentages and Trends • Descriptive Statistics • Making Management Decisions • Your.
Management Math for Libraries An Infopeople Workshop Jeanne Goodrich [email protected] Winter 2004-05 AGENDA • Numbers Issues • Percentages and Trends • Descriptive Statistics • Making Management Decisions • Your Action Plan What Is It? Math anxiety is a feeling of intense frustration or helplessness about one’s ability to do math. Math Anxiety • Primarily an American phenomenon • But, not a new one • Real or a myth? • Parents pass the phobia on to their kids • Math phobia takes hold gradually Math Proficiency Declines • 32% of fourth graders are proficient • 29% of eighth graders Percent Proficiency • 17% of 12th graders 35 30 25 20 15 10 5 0 4th grade 8th grade 12th grade We’ve Been Set Up! Until pulled from the market, Teen Barbie was programmed to say “Math is hard!” Those who supervise, plan, and manage in fields such as social work, librarianship, retail sales, school administration, and even publishing now need to be familiar with or at least willing to learn more math. Dealing With Math Anxiety • Recognize as emotional response (often learned) – unconstructive ways of addressing • rationalization • suppression • denial – constructive way to address • identify sources • accept the feeling without self-criticism • learn strategies Strategies for Approaching Math Problems • Define the problem, the number and operations to be used • Use smaller numbers • Examine related but easier problems or related but more general problems • Perform calculations in head, with paper, pencil or calculator • Work backwards from the solution • Draw pictures and diagrams Issues in Library Math • How are we doing? • How do we compare to others? – How much more or less are we doing? • How much does it cost to do what we do? – How much could we save if we didn’t do it? – How much do alternatives cost? • How do I interpret this report? • What else?? Percentages • Percentages show – parts of a whole (1/4 or 25%) – part of a group (1 in 4) – a ratio (1:4) – result of division (4 divided into 1) 1 4 Fractions and Percents You can convert between fractions and percents: 3 = .15 or 15% Divide 3 by 20 on your 20 calculator 36% = 36 = 9 x 4 = 9 100 25 x 4 25 Reduce to lowest terms Translating Percent Problems Percent problems are like story problems…you have to translate the problem statement into a math problem. 30% of what number is 16? Translation: 30% of • = 16 This involves solving for the unknown. Uh, Oh…algebra! Algebra is basically an approach to problem solving which frequently uses letters to represent numbers and performs arithmetic operations on them. Solving for X • Equation – a mathematical statement that tells us that one side is equal to the other: 3+4=7 • If you do something to one side, you have to do it to the other, to keep them equal: 3+4+3=7+3 • Holds true for subtraction, multiplication and division, also. 3+4=7 5 5 How it Works If x - 6 = 2, how much is x? To isolate 6, add 6 to both sides of the equation: x – 6 + 6 = 2 + 6 x=8 Check it out: 8 – 6 = 2 If 2x = 50, how much is x? 2x = 50 2 2 x = 25 If x = 9, how much is x? 2 x*2=9*2 x = 18 If 2x + 5 = 10, how much is x? 2x = 5 x=5 2 x = 2.5 An Application… Branch circulation at the Tree County Library totaled 5,397,892 last fiscal year. The circulation at the Main Library was one third of the total circulation. What was the total circulation for the system? 5,397,892 = 2x 3 5,397,892 x 3 = 2x 16,193,676 = 2x x = 8,096,838 Some Typical Questions • What is 25 percent of 24? • 6 is what percent of 24? • 6 is 25 percent of what? Displaying Percents Graphically Pie Charts • Show segments of a whole • Present only one variable Trends • Trends show where we’re going and where we’ve been • In our culture, upwards, left to right is generally good • Develop a time series (longitudinal) by tracking statistics at regular intervals over a period of time Circulation Trends Total Circulation 9000000 8000000 7000000 6000000 5000000 4000000 3000000 2000000 1000000 0 FY 85/86 FY 86/87 FY 87/88 FY 88/89 FY 89/90 FY 90/91 FY 91/92 FY 92/93 FY 93/94 FY 94/95 FY 95/96 Circulation/Expenditures Comparison 25000000 20000000 15000000 Total Circulation Operating Expenditures 10000000 5000000 0 FY 85/86 FY 86/87 FY 87/88 FY 88/89 FY 89/90 FY 90/91 FY 91/92 FY 92/93 FY 93/94 FY 94/95 FY 95/96 Circulation/Expense Comparison 9,000,000 25000000 8,000,000 20000000 7,000,000 6,000,000 15000000 5,000,000 4,000,000 10000000 3,000,000 2,000,000 5000000 1,000,000 0 0 FY FY FY FY FY FY FY FY FY FY FY 85/86 86/87 87/88 88/89 89/90 90/91 91/92 92/93 93/94 94/95 95/96 Total Circulation Operating Expenditures Trends and Forecasting • Don’t just put a trend line on your historical data and run it into infinity • Consider internal and external influences • budget • politics • changes in population • new facilities • changes in services, hours, collection • user habits and expectations Descriptive Measures for Interpretation and Analysis • Distributions – averages – spread – shape • Measure of variability – percentages – standard deviation Averages – Measures of Central Tendency • Mean – arithmetic average, sum of all numbers divided by the number of numbers in the series • Median – number in the middle of a data series • Mode – the number occurring most often Picking an Average to Use • Mean – sensitive to extreme values • Median – more representative of a range of values, particularly if great variance • Mode – describes relative popularity of specific values; only appropriate measure for categorical values Spread or Dispersion • Range – gives you a picture of the top and bottom of your data, lets you see how it might be skewed – Average becomes more meaningful with this information • you may want to lop off the outliers (extremely high or low values) – Can divide up into quartiles, deciles, or percentiles (quarters, tens and hundreds) Quartiles Example 2,500,000 Bottom 25% Top 75% 1,875,000 1,250,000 625,000 L C L TJ H S L T O P B D L M ID D O C RO NW S GR HL SE HG NP CA AL WO BE GS M HW / IR A F Using Percentiles Measurement Population of service area Total operating expenditures Operating expenditures per capita Circulation per capita Total Circulation Percentile for Actual Value Tree County PL 85th percentile 315,418 70th percentile $4,303,242 15th percentile $ 13.64 16th percentile 3.47 71st percentile 1,094,744 2d Approach to Spread Standard Deviation – a better measure – Deviation is the distance between any measurement in a set of data and the mean of the set. Standard deviation summarizes the degree to which the numbers in the distribution differ from the mean. – Rule of thumb: two thirds of a distribution is enclosed within one standard deviation, 95% within two, nearly everything within 3 Histogram Histogram 9 Positive Skew, skewed to the right 8 6 5 4 3 2 1 M or e 15 59 96 1. 75 11 43 95 3. 5 72 79 45 .2 5 0 31 19 37 Frequency 7 Determining the Z Score You can compute how many standard deviations above or below the mean a value is: z = (x - ) / x = specific element in the data set = the mean = the standard deviation Why Does This Matter? • Averages define a point around which other values tend to cluster • Measures of variability indicate how widely the values are dispersed around the average – Range shows difference between highest and lowest – Variance measures amount of dispersion in a distribution – Standard deviation measures dispersion of scores around the mean Correlation • Measures the strength of the relationship between two sets of data (two variables) • Will always be between –1 and +1 • Positive = when value of one increases, so does the other • Negative = when value of one increases the other decreases • The nearer r is to zero, the weaker the correlation Correlation Example Population/Operating Budget $20,000,000 $18,000,000 $16,000,000 Operating Budget $14,000,000 $12,000,000 $10,000,000 $8,000,000 $6,000,000 $4,000,000 $2,000,000 $0 0 50,000 100,000 150,000 Population 200,000 250,000 300,000 Turnover Rate/Coll Size 25. 0 Turnover Rate 20. 0 15. 0 10. 0 5. 0 0. 0 0 20, 000 40, 000 60, 000 80, 000 100, 000 120, 000 Collection Size 140, 000 160, 000 180, 000 Scatter Plots Important Warning! • Correlation is NOT the same as cause and effect • Cause and effect is tricky stuff: – A causes B – B causes A – C causes A and B – Chance Data Gathering and Analysis Why do it? • Measure and improve library services • Shape policy • Set goals and performance objectives • Formulate strategic goals and objectives • Justify existing or new services Sampling • Sampling is the process whereby items are selected from a population • The items are then analyzed in order to generalize the findings to the population • Samples should be representative of the population • Need to collect enough to accurately reflect full population Why Sample? • Less time consuming • Less costly • Some populations very large • Some inaccessible • Trial periods, pilots are forms of samples Possibilities • Work processes – check-ins per hour – items processed per hour – items shelved per hour • Equipment – uptime of public pcs – temperature throughout the building • Customer attitudes, satisfaction, use of services – hours of service Sample Design • Nonprobability neither known nor equal to probability – convenience – haphazard – judgment – purposive – quota – systemic • Easier, less costly • Probability each element in population has equal chance of inclusion – random – stratified • Takes more time, more accurate Sample Size Complicated formula boils down to: – 10% of population unless very large – No more than 1000 needed, regardless of population size – Very small samples likely to have more errors and be less representative Key Concepts • Confidence interval – range around which actual value for the population is likely to fall. Stated +/- 5, for example 43% of those surveyed indicated they would support the measure = 38-48% of the population • Confidence level – Tells you how sure you can be, how often the true percentage of the population would pick an answer within the confidence interval. 95% usually used. • Statistical significance – probably true, small probability of happening just by chance Summarizing Categorical Data • Categorical data captures qualities or characteristics (gender, age, race, ethnicity, language spoken at home) • Often summarized by reporting percentages in various categories • Cross Tabs or Two-way tables summarize information from two categorical variables at once, such as age and language spoken at home Regression Analysis A way of defining the extent to which two variables are related. Regression can be used in an attempt to predict things—but it can be tricky. For example, you may have a community survey and find, through regression analysis (that your professional surveyor does for you) that your Internet policies on filtering are a convincing reason to vote “No” to 25% of respondents Make Your Data Understandable • You’re not alone in not liking numbers • Dense text obfuscates…as do unfamiliar words • Think about your audience when choosing how to present your data • Use charts and graphics • Be sure you make the point you intend to make • Aggregate and disaggregate for your reader This LOOKS like more, but is really less OCLC report: Libraries: How They Stack Up www.OCLC.org/index/compare/ Aggregation & Disaggregation • “Roll up” data to make it more understandable – individual branch, department or unit statistics into a summary – monthly into yearly – five and ten year summaries of key statistics • “Drill down” to get more detail – identify variables – identify data errors or collection errors library material Video Multnomah County Library, FY 04 Compiled from holdings and circulation reports; turnover computed. Clear information from several reports. circulation holdings turnover (dynix only) (excl. Lost) rate 1,934,571 75,639 25.6 DVD 816,935 35,731 22.9 CD-ROM 109,696 4,854 22.6 2,052,166 114,326 18.0 Audio Board books & catalogued Pb 748,298 45,663 16.4 334,186 21,116 15.8 Juvenile Fiction 651,668 61,837 10.5 Picture books (JE) 1,615,009 154,223 10.5 Fiction 2,098,367 207,922 10.1 464,088 46,541 10.0 Non-fiction 6,018,557 673,112 8.9 Juvenile Non-fiction 1,046,573 142,770 7.3 Large Print 165,993 24,653 6.7 Int'l Languages 236,376 39,363 6.0 Music Scores 160,538 44,281 3.6 70,026 247,646 0.3 Total 18,523,047 1,939,677 9.5 CD Young Adult Other Correlation of FTEs and Circ C o r r el at i o n of FTE to C i r i s . 52 2 800,000 700,000 Circulation 600,000 500,000 400,000 300,000 200,000 100,000 0.000 10.000 20.000 30.000 40.000 50.000 FTEs Financial Analysis What financial problems have you dealt with or would you like to be better equipped to deal with? Are there aspects of your budget process that you’d like to understand better? Ratios, Rates and Percentages • Ratio is a fraction that divides two quantities: 3 girls to 2 boys; 1 manager to 12 staff members • A rate is a ratio that reflects some quantity per a certain unit: circulation is 4.5 per capita, turnover rate is 9.8 per Adult Fiction title • A percentage reflect a proportion of the whole: 15% of the materials budget goes for large type materials Cost Analysis • Cost Allocation – How would you approach allocating overhead costs to library programs? • Cost per Unit – Can you relate outputs to inputs? Do you know what it costs you to catalog a book? A set of books on tape? What variables are important? Cost Benefit Analysis A decision-making technique to develop quantitative information on how to allocate resources. cost-benefit ratio: estimated benefits costs Objective: benefits outweigh costs Break Even Analysis The point where total costs equal total revenues Total fixed costs = Break even point Contribution margin If a library buys 300 book bags from a supplier for $12 each and plans to sell them for $15, how many do they have to sell to break even? Return on Investment Business concept you might be able to apply in the library Rate of return = revenue cost center profit investment Library gift shop: $35,000 = 30% $117,500 Your Action Plan • What project will you take on using what you’ve learned today? • Have you found some ways to answer your questions with numbers? • Have you found some new ways to use numbers that you hadn’t known of before today? • What are you going to do as you prepare your next budget? Thank You! [email protected]