Bus 7471 - Analytics Flashcards
From the book Keeping Up with the Quants what are the steps to conducting HR analytic projects? (hint: 3 main ideas and then each have components under them)
*mentioned in every class
Framing the problem
- problem recognition:
- this means linking HR with wider business issues
- Using financial reports, ee surveys, external benchmark data
- Example of outcomes org agility and resilience, culture, productivity and performance
- Using HR data to link to the business impacts
- Review of previous findings
Solving the problem
- modelling
- Data collection
- Data analysis
Communicating and action on the results
- Results presentation
Give me a definition of what HR analytics means to you? Can you give some examples of what it is?
(hint there are two definitions and one is with that funnel)
Class 2
“An HR practice enabled by information technology that uses descriptive, visual and statistical analyses of data related to HR processes, human capital, org performance and external economic benchmarks to establish business impact and enable data driven decision making”
Another description:
You mix
- data & analysis
- people questions
- technology
=
You get insights, decisions, actions and impact
Other ways to define it:
- an emerging trend within HRM
- A strategic business process
- A competitive advantage
- A game changer
Examples
- workforce planning
- gender pay gap
- employee turnover
- probation rates
- number of hires based on referrals
(can go on and on)
Describe the 6 steps to an evidence-based management approach to management?
Definition of evidence-based management is about management using evidence based practices to make decisions through the conscientious, explicit and judicious use of the best available evidence form multiple sources.
Hint: they all start with the letter A
Class 2
So management will:
- Asking then they will
- Acquiring followed by
- Appraising
- Aggregating
- Applying
- Assessing- at the end of this process there is an increase likelihood of a favourable outcome
The type of data you have will determine what?
Class 2
The data type will tell me what analysis I will use; will I use descriptive, predictive or prescriptive.
What are the 3 types of HR analytics?
Class 2
- Prescriptive
- applies mathematical and computational sciences to suggest decision options to take advantage of the results of descriptive and predictive analytics. It specifies both the actions necessary to achieve predicted outcomes and the interrelated effects of each decision.
- takes both predictive and descriptive results and uses it to identity where they sit on a table
- Y axis of table is the impact (regression and standardized coefficient)
- x axis of table is presence (descriptive and mean value)
- Then where it sits on the table will determine if you need to maintain (right upper and right lower), improve (lower left) or correct (upper left). - Predictive
- Modelling/hypotheses; X predictors -> Y outcomes
- Correlation analysis is used to measure strength of the association (linear relationship) between 2 variables;
- Correlation range is -1 and 1.
*there is another card with more information
- shopping list is called impact
- correlation is the space between input and output; give “kind of” if they are together or apart; can’t be used as a causation prediction to outcomes;
- regression coefficient; true story need to get description so first step is correlation?? - Descriptive (2 types)
- good for understand historical data, informed decision-making, identifying key metrics, resource allocation, performance tracking and reporting
CON - Continuous variable:
- Mean; look at average
- Examples “how old are you”/age, job satisfaction
- 1 means 1 and 2 means 2; means the higher, longer, happier the number
- most of Likert scales are treated as continuous
CAT - Categorical variable:
- Frequency %
- Examples “which age group are you in”, gender and job grade, location, education, department, country’s, nationalities
- 1 does not mean 1 and 2 does not mean 2; so gender for female is 1 and male is 2. Number doesn’t mean that larger the better or smaller the worse. The difference in numbers does not give the meaning
What are 4 issues with data?
Hint RTCA
Class 2
- Reliable
- Timely/real time
- Consistent
- Accurate
List 5 dark sides of HR Analytics with an example if possible
Class 3
- Bring about an illusion of control and reductionism (Microsoft MyAnalytics tool)
- Lead to estimated predictions and self-fulfiling prophecies (police using historical crime data leading to discrimination)
- Foster path dependencies (Amazon screening system)
- Impair accountability and transparency (if you don’t understand how AI is making decisions could get risky if using to choose who to layoff)
- Marginalize human reasoning and erode managerial competence (online labor market system which led to ee not understanding evaluation criteria)
List best practices to reduce AI Bias
Class 3
- Business organizations and leaders will need to stay up to-date on this fastmoving field of research
- When your business or organization is deploying AI, establish responsible processes that can mitigate bias.
Google AI has published recommended practicwe
IBM has a “Fairness 360” framework that pulls together common technical tools
- Engage in fact-based conversations around potential human biases.
- Invest more, provide more data, and take a multi-disciplinary approach in bias research (while respecting privacy) to continue advancing this field
- Invest more in diversifying the AI field itself. A more diverse AI community would be better equipped to anticipate, review, and spot bias and engage communities affected.
The second type of data analytics is predictive.
What does this do?
Give some examples.
Depending on the number you get what does it mean?
Class 4
- Modelling/hypotheses; X predictors -> Y outcomes
- X are independent variable, predictor, observable variable, input variable; leadership, training & dev oppo, org design, fairness in PM, fairness in RM, HRM practices
- Y are dependent variable, outcome, predicted variable, output variable; job satis, wellbeing, engagement, productivity, ROI, turnover intention, turnover, creativity
- Correlation before regression. Correlation does not mean causation. Correlation analysis is used to measure strength of the association (linear relationship) between 2 variables; no causal effect is implied;
- Correlation range is -1 and 1:
-1 means the stronger the negative linear relationship
1 means the stronger the positive linear relationship/stronger the correlation between 2 variables is (too high is not good if you have r > .50 may need to think are they measuring the same thing or not (validity)
0 means the weaker the linear relationship
What is digital twin transition?
EU Aims: sustainable, fair, competitive and inclusive
Green: achieve sustainability, and combat climate change and environmental degradation
Digital: harness digital technologies for sustainability and prosperity and empower citizens and business
Digital technologies can be key enablers for reaching the European Green Deal objectives
Successfully managing the green and digital ‘twin’ transitions i sthe cornerstone for delivering a sustainable, fair and competitive future
Examples of sustainable digital tech: AI, smart robots, data-driven tech, internet of things, computing infrastructure, comm tech, software and service tech, extended reality and metaverses
