Abstract
Data-driven decision-making and accompanying data analytic skills are increasingly important. Data analytics involves several steps: identifying the study's objective, determining the relevant data to collect, choosing the appropriate instruments to collect those data, analyzing that data, recommending appropriate actions, implementing them, and evaluating the implementation. Appropriately applying and analyzing statistics can be daunting, even for managing leaders. This chapter provides guidance in preparing data, matching the data with the appropriate statistical test, and making sense of the statistics. Library-based case studies illustrate how data analytics enables leaders to control and improve library processes.
TopIntroduction
Today’s leaders need to use data effectively in order to act strategically. Considering all the data available about organizations, one might think that data-driven decision-making would be easy. However, several steps are involved in data analytics: identifying the study’s objective, determining the relevant data to collect, choosing the appropriate instruments to collect those data, analyzing that data, recommending appropriate actions, implementing them, and evaluating the implementation.
One of the main sticking points in such data procedures appears to be the analysis itself. How do leaders make sense of the data? What do the story behind the data tell? While accurately representing the data is a good first step, explaining the data and making inferences about the findings is necessary in order to make logical, feasible recommendations for action.
Several factors may account for this phenomenon: management’s lack of statistical background, statistician’s lack of organizational and management knowledge, lack of technical or statistical personnel, differing status between management and technical (e.g., statistician) personnel, perceived disconnect between data and management practices, perceived lack of available time, perceived poor ROI (return on investment) of data analytics, poor data quality, lack of integration of data with existing infrastructure, insufficient systems to process data efficiently, perceived lack of data analytics tools (or lack of knowledge about such products), other higher-priority management demands (Finos, 2015; Henschen, 2014). Because of these identified needs, quantitative data analytics practices are detailed in this chapter, focusing on quality control. Libraries constitute the organizational context.
Key Terms in this Chapter
Variable: factor (e.g., staff, hours of operation, circulation of materials, number of computers).
Descriptive Statistics: Common statistics include minimum and maximum, mean (average), median (middle value), quartiles, and standard deviations (which measure variability).
Non-Parametric Statistics: Statistics that do not rely on data belonging to any particular distribution.
Quantitative Data: Numerical data.
Normal Distribution: A data set in which most values cluster in the middle of the range and the rest tape off symmetrically toward both ends.
Qualitative Data: Non-numerical data (e.g., interviews, videos, artifacts).
Data Analytics: A scientific and systematic approach to examine raw data in order to draw valid conclusions about them. Data are extracted and structured, and qualitative and quantitative techniques are used to identify and analyze patterns.
Sampling: Taking a predetermined number of observations (responses) from a larger population.
Quality Control: Procedures intended to ensure that a product or service meets a defined set of quality criteria.