HRA-Chap 1 (2020)

HR ANALYTICS-CHAPTER 1
Human Resource Analytics

1
(18B12HS412)
Course Objective
• This Course will introduce the concept and
applications of human resource analytics. The
students will develop understanding of role and
importance of analytics in making effective human
resource decisions varying from hiring, training,
performance appraisal, engagement, retention,
compensation and others. The course will teach
students the methods, principles and techniques for
conducting human resource analytics.

2
(18B12HS412)
Course Outcomes
C401- Gain an understanding of the different analytical techniques used for solving
20.1 HR related problems.
C401- Apply descriptive and predictive analysis techniques to understand trends and
20.2 indicators in human resource data.
Analyze key issues related to Workforce Planning, Diversity, Talent Sourcing,
C401-
Talent Acquisition, Talent Engagement, Training and Intervention,
20.3
Performance Management and Retention using analytical techniques.
C401- Critically asses and evaluate the outputs obtained from analytical tools and
20.4 recommend HR related decisions.
C401- Create and hypothesize the situation given and check for its validity using
20.5 appropriate analytical technique.

3
(18B12HS412)
Module No. Subtitle of the Module Topics in the module No. of Lectures
for the module
1. Introduction to Human Understanding the need for mastering and 8
Resource (HR) Analytics utilizing HR analytic techniques, Human
capital data storage and ‘big (HR) data’
manipulation, Predictors, prediction and
predictive modeling, Current state of HR
analytic professional and academic
training, HR’s Contribution to Business
Value, the Changing Nature of HR.
2. Human Resource information Understanding HR metrics and data, Data 8

systems and data collection, tracking, entry, Data availability
in the entire Employment Lifecycle,
Approaches and costs of collecting HR
related data, Analysis software options,
Using SPSS, Preparing the data
3. Analysis Strategies From descriptive reports to predictive 10
analytics, Statistical significance, Data
integrity, Types of data, Categorical
variable types, Continuous variable types,
Using group/team-level or individual-level
data, Dependent variables and independent
variables, Introduction of tools for HR data
analysis: Correlation, Regression, Factor
Analysis, Cluster Analysis, Structural
equation modeling.

4
(18B12HS412)
Module No. Subtitle of the Module Topics in the module No. of Lectures
for the module
4. Application of Human Workforce Planning Analytics, Diversity 10
Resource Analytics Analytics, Talent Sourcing Analytics,
Talent Acquisition Analytics, Talent
Engagement Analytics, Training and
Intervention Analytics, Analytical
Performance Management, Retention
Analytics.
5. Future of Human Resource Rise of Employee Behavioral Data, 4
Analytics Automated Big Data Analytics, Big Data
Empowering Employee Development,
Quantification of HR, Artificial
Intelligence in HR.
Total number of Lectures and Tutorials 40

5
(18B12HS412)
What Is Human Resource
Management (HRM) ?
• The comprehensive set of managerial
activities and tasks concerned with developing
and maintaining a qualified workforce—
human resources—in ways
that contribute to organizational effectiveness.

6
(18B12HS412)
Human Resource Management as a Center for Expertise

1–7
(18B12HS412)
What is HR Analytics
• The role of HR – through the management of an organization’s
human capital assets – is to impact four principal outcomes:
(a) generating revenue, (b) minimizing expenses, (c) mitigating
risks, and (d) executing strategic plans.
• “HR analytics is a methodology for creating insights on how

investments in human capital assets contribute to the success
of those four outcomes. This is done by applying statistical
methods to integrated HR, talent management, financial, and
operational data.”

8
(18B12HS412)
HR analytics is also often referred to as:
• People analytics
• Talent analytics
• Workforce analytics
The below graph provided by Google Trends shows search
interest for these terms since 2004. Both the terms HR
analytics and people analytics have grown in popularity and
continue to gain interest.

9
(18B12HS412)
How Does HR Analytics Drive
Business Value?
HR has access to valuable employee data. How can this
data be used to enable change in the organization?
There is a great deal of discussion on replicating the

consumer experience in the employee experience.
Essentially, the data pertaining to consumer behavior and
mindset can inform strategies to maximize sales by
capitalizing on those factors. Similarly, the data that
informs the HR function can be used to improve the
employee experience, and in turn, maximize business
outcomes.

10
(18B12HS412)
What Metrics Does HR Analytics Measure?

11
(18B12HS412)
1. Time to hire (time in days)
An important metric for recruitment is the ‘time to hire’. This is the number of days
between a position opening up and a candidate signing the job contract. It’s an excellent
way to measure the efficiency of the recruitment process and provides insight into the
difficulty of filling a certain job position.
2. Cost per hire (total cost of hiring/the number of new hires)

Like the time to hire, the ‘cost per hire’ metric shows how much it costs the company to
hire new employees. This also serves as an indicator of the efficiency of the recruitment
process.
3. Early turnover (percentage of recruits leaving in the first year)

This is arguably the most important metric to determine hiring success in a company.
This early leaver metric indicates whether there is a mismatch between the person and the
company or between the person and his/her position. Early turnover is also very
expensive.
4. Time since last promotion (avg time in months since last internal promotion)
This rather straightforward metric is useful in explaining why your high potentials leave.
5. Revenue per employee (revenue/total number of employees)

This metric shows the efficiency of the organization as a whole. The ‘revenue per
employee’ metric is an indicator of the quality of hired employees.
(18B12HS412)
12
6. Performance and potential (the 9-box grid)
The 9-box grid appears when measuring and mapping both an individual’s
performance and potential in three levels. This model shows which employees are
underperformers, valued specialists, emerging potentials or top talents.
7. Billable hours per employee

This is the most concrete example of a performance measure, and it is especially
relevant in professional service firms (e.g. law and consultancy firms). Relating this
kind of performance to employee engagement or other input metrics makes for an
interesting analysis.
8. Engagement rating
An engaged workforce is a productive workforce. Engagement might be the most
important ‘soft’ HR outcome. People who like their job and who are proud of their
company are generally more engaged, even if the work environment is stressful and
pressure is high.
9. Cost of HR per employee (e.g. $ 600)

This metric shows the cost efficiency of HR expressed in dollars.
10. Ratio of HR professionals to employees (e.g. 1:60)

Another measure that shows HR’s cost efficiency. An organization with fully
developed analytical capabilities should be able to have a smaller number of HR
professionals do more. Human Resource Analytics
13
(18B12HS412)
11. Ratio of HR business partners per employee (e.g. 1:80)
A similar metric to the previous one. Again, a set of highly developed analytics
capabilities will enable HR to measure and predict the impact of HR policies. This will
enable HR to be more efficient and reduce the number of business partners.
12. Turnover (number of leavers/total population in the organization)

This metric shows how many workers leave the company in a given year. When
combined with, for instance, a performance metric, the ‘turnover’ metric can track the
difference in attrition in high and low performers.
13. Effectiveness of HR software

This is a more complex metric. Effectiveness of, for instance, learning and development
software are measured in the number of active users, average time on the platform,
session length, total time on platform per user per month, screen flow, and software
retention. These metrics enable HR to determine what works for the employees and
what does not.
14. Absenteeism (absence percentage)

Like turnover, absenteeism is also a strong indicator of dissatisfaction and a predictor of
turnover. This metric can give information to prevent this kind of leave, as long-term
absence can be very costly.
14
(18B12HS412)
15
(18B12HS412)
16
(18B12HS412)
SOME COMMON
APPLICATIONS

17
(18B12HS412)
RETENTION
The cost to replace an employee could be over 200% of their annual salary, according to
AmericanProgress.org. The true cost might even be higher due to training/onboarding,
lost productivity, recruitment, and decreased morale among other employees. Losing an
employee that’s in the top 1% of performers could mean the difference between growth
and decline. For this reason, decreased attrition and improved employee engagement
are often top priorities for HR departments. HR analytics can help improve retention
through a churn analysis that looks at data points like:
•Current churn rate

•Attrition by department
•Attrition by estimated commute time
•Similar attributes of employees with longer tenure
•Similar attributes of employees who leave within 1 year
•Onboarding experience
•Survey data
•Qualitative data such as employee interviews
•Employee performance data to forecast future attrition

18
(18B12HS412)
Through this data-driven approach,
HR analytics can illuminate the
major causes of attrition, and new
policies, along with training
programs, can be put in place to help
mitigate the problem. For example,
data might show that high-aspiration
employees are not challenged or
employees are frustrated with a
certain management style. Human
resources analysis will reveal these
issues, and then it will be up to
leadership to act. It’s also possible to
spot an at-risk employee before they
leave so preemptive actions can be
taken to resolve issues. For example,
a once high-performer may not be as
productive because he feels he or she
is underpaid. An analysis of
productivity alongside a comparison
of market-value salaries can help
spot this.
19
(18B12HS412)
EMPLOYEE PERFORMANCE
As qualified candidates become more difficult to find and retain,

improvements in performance evaluation can prove to be tremendously
valuable. HR analytics leverages your employee data to determine who
your best and worst performers are, based on factors like past work
experience, length of employment, and to whom they report. Common
characteristics between groups of employees can emerge, which can be
used to properly motivate employees and ensure they receive the right
rewards. Career progression is an important component to monitor, as
well, since employees are much more likely to leave if they’re not
satisfied with their career trajectory. Analysis of promotion rates,
promotion wait times, and qualitative data from employees will help
HR meet employee expectations and keep them motivated. Leadership
traits may emerge in some employees, and companies can determine
which characteristics correlate with future leaders. HR can then foster
this potential, resulting in long-term ROI for the organization.

20
(18B12HS412)
RECRUITING
Organizations need to first determine who their ideal candidate is and what skills
are needed for the job. This can be discovered through discussions with hiring
managers, analysis of previously successful hires, and available market data.
Organizations can then match their need to where the largest pool of applicants
with that skillset resides. For example, the best cities for PR jobs according to
Forbes are Los Angeles, New York, and Washington D.C., so a recruiter may want
to focus efforts in these cities if they aren’t able to find local talent for the job.
Likewise, a recruiter may look at college graduate data to see where the best new
talent is if they need to fill an entry-level position. Talent acquisition teams also
need to adopt channel optimization strategies based on recruitment analytics. A
continuous process of evaluating which channels drive the best candidates at the
most efficient price is extremely important. Data analytics may reveal one source
has a lower cost per applicant, while another may provide the best talent. HR
analytics can also be used to craft the best recruitment message to the right
candidate based on internal, open-source, and third-party data. This is all data that
HR analytics platforms can consume and transform into easily digestible formats
such as dashboards and charts. Recruitment relies heavily on metrics gathered
during the interview process.

21
(18B12HS412)
Important metrics include:
•Average number of applicants
•Number of applications to offer
•Scheduled maintenance windows for
system patching, security updates, and
AWS infrastructure management
•Number of final rounds to offer
•Offer acceptance rate
Over time, HR analysts can also see

which candidate characteristics predict
the best employees. Pre-screen exams
are often used to test candidates, and
analytics can help determine what types
of questions these exams should
contain. For example, a candidate who
applied for a Python developer role can
be asked a series of questions to assess
competence within the program.
22
(18B12HS412)
EMPLOYEE DEVELOPMENT
Almost 40% of American employers say
they cannot find people with the necessary
skills for entry-level jobs and almost 60%
say entry-level candidates are not prepared
professionally.
HR analytics is also playing an increased

role in evaluation of employee
development (or workforce development)
programs to close the skills gap. Analytics
tools can help human resource
management assess company needs,
allocate resources to train employees most
likely to fill those needs, and then evaluate
the results. The goal is to close the skills
gap so a company can compete with
increasingly agile global competitors.
23
(18B12HS412)
Data visualizations fed by employee and customer feedback data can show the
evolution of employee professional growth. For example, a retail clothing
chain may discover through customer feedback that sales associates are
friendly, but have trouble providing excellent customer service when items are
not in stock. An employee development program can be put in place to educate
associates on the keys to customer success. This could take the form of a pilot
training program for employees in the lowest-performing locations.
The best employee development and talent management programs align

employee goals with business goals, and investment is often dependent on the
ability to tie these to revenue. The process is successful when ROI is achieved,
creating a feedback cycle where success increases revenue and investment in
development.

24
(18B12HS412)
WORKFORCE PLANNING
Not only do organizations need to ensure they have employees with the
right skills presently, but they also need to be able to anticipate future
departures or changes in the organization’s needs, particularly for business
growth. HR needs a data-driven approach to successfully navigate the
ever-changing composition of organizational talent. This ensures that
workforce and capacity planning are optimized.
For example, an electronics OEM may need to anticipate a greater need for
call center support with the release of new products in the next quarter.
Once again, HR needs to be aligned with the goals of the business and
upcoming changes. Predictive analytics and data analytics platforms are
changing the way HR can utilize and anticipate these changes.

25
(18B12HS412)
EMPLOYEE ENGAGEMENT
An engaged workforce is critical to attracting and engaging talent.
The challenge is determining which factors will result in revenue for
the company. This makes employee engagement a very difficult
metric to quantify and take action on.
•How can HR analytics help with employee engagement?
•Statistical analysis of employee engagement survey data
•Discriminant analysis to identify the needs of different segments
within the organization
•Factor analysis to correlate engagement initiatives with retention
and productivity
•Inform testing of new ideas to measure lift in engagement
Google is a perfect case study in how people analytics can be used

to increase employee engagement. They’re one of the most data-
driven cultures in the world. Its People Operations team utilizes a
mix of quantitative and qualitative data to measure what employees
value most and to keep them engaged.
26
(18B12HS412)
They discovered, through a combination of employee surveys and
productivity analytics, that great managers tend to have the following
qualities:
1. Coaching skills
2. Does not micromanage
3. Is genuinely concerned for well-being of employees and expresses
this
4. Results-oriented
5. Excellent communicator (listens and shares)
6. Develops employees
7. Clear vision and strategy for team
8. Possesses key technical skills to help guide team

27
(18B12HS412)
COMPENSATION AND INCENTIVE
PROGRAMS
Compensation is often the largest business expense, which underlines its
importance in organization decision-making. Data analytics platforms can help
analyze large volumes of employee and market data to achieve a competitive
advantage. Smart HR analysts will keep track of what competitors offer
employees to ensure top talent is attracted to the company. They can also look
at exit surveys during the recruiting process and declined offers to better
understand the compensation and incentive landscape. HR analytics can
provide:
Incentive programs to motivate employees to maximize productivity. Data

analysis can provide insight into the most impactful monetary and non-
monetary rewards. For example, an HR department might implement a referral
program where employees receive a bonus for referring new hires. With
enough data, HR can determine what the ideal bonus amount is and cross-
reference this to the increase or decrease in quality of hires.

28
(18B12HS412)
Sales team motivation is another application. What quota should be assigned
to each salesperson and what should their bonus structure look like? This
may vary on the team, region, and products sold.
Executive compensation analysis needed to attract and retain top leaders.

HR analytics can analyze the market rates for executives at similar
companies, bonuses that properly motivate, and perks to provide. This is
critical to the direction of the organization and can either make or break it.

29
(18B12HS412)
WHAT ARE THE BENEFITS OF HR
ANALYTICS?
In sum, HR analytics will move from an operational partner to a more
strategic center of excellence. Companies are now realizing company
success is built on people, and HR analytics can light the way from
intangible theory-based decisions to real ROI through the following:
•Better hiring practices

•Decreased retention
•Task automation
•Process improvement
•Improved employee experience
•More productive workforce
•Improved workforce planning through informed talent development

30
(18B12HS412)
WHAT ARE THE BIGGEST CHALLENGES OF
HR ANALYTICS?
The road to actionable HR analytics is not always easy. There are several
challenges organizations need to overcome so they can reap the rewards:
•Finding people with the right skill set to gather, manage, and report on the data
•Data cleansing
•Data quality
•Too much data to parse or not knowing what data is most important
•Data privacy and compliance
•Proving its worth to executive leadership
•Tying actions and insight to ROI
•Identifying the best HR technologies to keep track of the data

31
(18B12HS412)
TOOLS FOR HR ANALYTICS

32
(18B12HS412)
33
(18B12HS412)
34
(18B12HS412)
35
(18B12HS412)
36
(18B12HS412)
SPSS is one of the most commonly used HR
analytics tools in social sciences. Thanks to its
user-friendly interface you’re able to analyze
data without having extensive statistical
knowledge. In addition, SPSS is often used
within the field of social science. This means
that a lot of HR professionals know how to use
it, especially the ones with an interest in data
analysis.
This is also the reason why we put SPSS on the

list and not its biggest competitor, SAS. SAS is
more widely used outside of the social science
field. However, SAS has a steeper learning
curve. In addition, SPSS shares many
similarities with Excel which makes it easier to
work with.

37
(18B12HS412)
R is the most used HR analytics tool. R is great
for statistical analysis and visualization which
is very suited to explore huge data sets. It
enables you to analyze and clean data sets with
millions of rows of data. In addition, it lets you
to visualize your data and analysis, like what
you see below.
RStudio is an open source and enterprise-ready

professional software package for R. It
basically does everything that R does, but has a
friendlier user interface. The interface contains
a code editor, the R console, an easily
accessible workspace, and history and room for
plots and files. You can take a look at an
example of this below.

38
(18B12HS412)
When we talk about HR analytics tools, we
shouldn’t forget the basics.
Excel is where most of us started. It’s no

surprise that when you manually extract data
form any of your HR systems, it most likely
comes in the form of a comma-separated value
(CSV) file. These files can easily be opened
and edited using Excel.
The good thing about Excel is that it’s very

intuitive to most of us HR data geeks and
therefore easy to use.
For example, if you wanted to check how

clean your data is, you can easily transform a
dataset into a table and check each column’s
data range for outliers.

39
(18B12HS412)
REFERENCES
Edwards and Edwards, Predictive HR Analytics. Mastering the HR Metric, Kogan
Page, Limited, 2019
Banerjee, Pandey and Gupta, Practical Applications of HR Analytics, Sage, 2019
Bhattacharyya, HR Analytics: Understanding Theories and Applications, Sage, 2017
Isson, Harriott and Jac Fitz-enz, People Analytics in the Era of Big Data: Changing
the Way You Attract, Acquire, Develop, and Retain Talent, Wiley, 2016
Guenole, Ferrar and Feinzig, The Power of People: How Successful Organizations
Use Workforce Analytics To Improve Business Performance, First Edition, Pearson,
2017
Sesil, Applying Advanced Analytics to HR Management Decisions: Methods for
Selection, Developing, Incentive and Improving Collaboration, Pearson, 2014

40
(18B12HS412)
REVIEW QUESTIONS
1. ________is one of the most 2. _________ is a tool used to clean
commonly used HR analytics tools in your data , you can easily transform a
social sciences. dataset into a table and check each
a)Google Analytics column’s data range for outliers.
b)SPSS a)Python
c)Python b)Excel
d)R-Studio c)MicroStrategy
d)Workday
3. _________ is an enterprise analytics 4. _________ is an open-source

and mobility platform which includes programming language for statistical
R, Python and Google Integration. computing and graphics.
a)Workday a)Google Analytics
b)R-Studio b)SPSS
c)MicroStrategy c)Python
d)SPSS d)R programming

41
(18B12HS412)
5.__________ is a high level programming 6._____________ is an important metric for
language made for general programming. recruitment which measures the number of days
a)Python between a position opening up and a candidate
b)Excel signing the job contract
c)MicroStrategy a)Cost per hire
d)Workday b)Time since last promotion
c)Time to hire
d)Revenue per employee
7.____________ metric is arguably the most 8._____________ metric shows the cost
important metric to determine hiring success in a efficiency of HR expressed in dollars.
company. a)Billable hours per employee
a)Early turnover b)Cost of HR per employee
b)Revenue per employee c)Performance and potential
c)Turnover d)Turnover
d)Absenteeism
9.____________metric shows how many workers 10. Like turnover__________ is also a strong
leave the company in a given year. indicator of dissatisfaction and a predictor of
a)Absenteeism turnover.
b)Turnover a)Turnover
c)Early turnover b)Effectiveness of HR software
d)Time since last promotion c)Absenteeism
d)Performance and potential

42
(18B12HS412)
11. The _______________ metric shows how 12. The _______________ metric is useful in
much it costs the company to hire new employees. explaining why your high potentials leave.
a)Cost per hire a)Early turnover
b)Time to hire b)Engagement rating
c)Time since last promotion c)Time since last promotion
d)Early turnover d)Revenue per employee
13.______________ is one of the leading HRIS 14.___________ metric shows the efficiency of
tools in the market due to its cloud-centric system the organization as a whole, it is an indicator of
architecture. the quality of hired employees.
a)Python a)Revenue per employee
b)Excel b)Ratio of HR professionals to employees
c)MicroStrategy c)Absenteeism
d)Workday d)Effectiveness of HR software
15. The most popular web analytics tool is

__________________ , it's a freemium tool that's
easy to implement and use.
a)Google Analytics
b)SPSS
c)Python
d)R programming
43
(18B12HS412)
DATA & ANALYSIS STRATEGIES
Human Resource Analytics (18B12HS412) 44

HR Research Tools and Techniques
• HR research can be either quantitative or qualitative.
• Quantitative data are gathered through; experimental/clinical trials,

observations and recording of events, collection of data from databases
such as enterprise resource planning (ERP) and management information
systems (MIS), survey reports and so on.
• Qualitative research covers historical data analysis and even a collection of

narrative data for understanding phenomena through discourse analysis.
• In all types of research such as descriptive, correlational, casual-

comparative and experimental, statistical analysis helps.

Types of Data Analysis
• Quantitative Methods
– Testing theories using numbers
• Qualitative Methods
– Testing theories using language
• Magazine articles/Interviews
• Conversations
• Newspapers
• Media broadcasts

The Research Process

Initial Observation
• Find something that needs explaining
– Observe the real world
– Read other research
• Test the concept: collect data
– Collect data to see whether your hunch is correct
– To do this you need to define variables
• Anything that can be measured and can differ across
entities or time.


Generating and Testing Theories
• Theories
– An hypothesized general principle or set of principles that
explain known findings about a topic and from which new
hypotheses can be generated.
• Hypothesis
– A prediction from a theory.
– E.g. the number of people turning up for a Big Brother
audition that have narcissistic personality disorder will be
higher than the general level (1%) in the population.
• Falsification
– The act of disproving a theory or hypothesis.


Measurement Error
• Measurement error
– The discrepancy between the actual value we’re
trying to measure, and the number we use to
represent that value.
• Example:
– You (in reality) weigh 80 kg.
– You stand on your bathroom scales and they say
83 kg.
– The measurement error is 3 kg.

Validity
• Whether an instrument measures what it set out
to measure.
• Content validity
– Evidence that the content of a test corresponds to
the content of the construct it was designed to cover
• Ecological validity
– Evidence that the results of a study, experiment or
test can be applied, and allow inferences, to real-
world conditions.

Reliability
• Reliability
– The ability of the measure to produce the same
results under the same conditions.
• Test-Retest Reliability
– The ability of a measure to produce consistent
results when the same entities are tested at two
different points in time.

How to Measure
• Correlational research:
– Observing what naturally goes on in the world without directly
interfering with it.
• Cross-sectional research:
– This term implies that data come from people at different age
points with different people representing each age point.
• Experimental research:
– One or more variable is systematically manipulated to see their
effect (alone or in combination) on an outcome variable.
– Statements can be made about cause and effect.

Levels of Measurement
• Categorical (entities are divided into distinct categories):
– Binary variable: There are only two categories
• e.g. dead or alive.
– Nominal variable: There are more than two categories
• e.g. whether someone is an omnivore, vegetarian, vegan, or fruitarian.
– Ordinal variable: The same as a nominal variable but the categories have a
logical order
• e.g. whether people got a fail, a pass, a merit or a distinction in their exam.
• Continuous (entities get a distinct score):
– Interval variable: Equal intervals on the variable represent equal differences
in the property being measured
• e.g. the difference between 6 and 8 is equivalent to the difference between 13 and 15.
– Ratio variable: The same as an interval variable, but the ratios of scores on
the scale must also make sense
• e.g. a score of 16 on an anxiety scale means that the person is, in reality, twice as
anxious as someone scoring 8.

• In HR research and decisions we make use of different levels of
measurements, i.e., nominal, ordinal, interval, and ratio.
• Nominal measurement is a classification or objects in two or more

categories; hence we call it categorical measurement.
• Ordinal measurement classifies objects in order from highest to

lowest, from the most to least. It can indicate one object is better
than the other, but cannot say, how better it is.
• Interval measurement combines both the characteristics of nominal

and ordinal measurements. Employees are measured using such
measurement tool, which can have a scale with arbitrary maximum
and an arbitrary minimum score (say zero point).
• Ratio measurements along with the properties of interval

measurement can analyse differences in scores and the relative
magnitude of scores. All these measurements require usage of
different scales, and then make use of statistics for better inferences.
Types of data
• Categorical variable types: A categorical variable is one that is made up of
categories. When you think about what this data looks like, there is no
immediate quantitative or numerical value.
• Example:
01 – Australia
02 – United Kingdom
03 – United States
04 – Hong Kong
05 – Singapore
06 – Canada
So although the system stores country as a number, it is not a numeric value with
which we could meaningfully add, subtract, multiply or divide.
• There are three different types of categorical variable: binary, nominal and
ordinal
Binary variable
• Binary variables are categorical variables that are unique in

that they name only two distinct entities or ideas.
Gender: male/female
Attended customer service training: yes/no

Nominal variable
• Nominal variables are categorical variables that name three or

more distinct entities. They are in no particular order and
each value is mutually exclusive.
Home office location: Australia, United Kingdom, Canada, United

States, Singapore, Hong Kong, etc.
Division: Sales, Legal, HR, Finance, etc.

Ordinal variables
• Ordinal data takes things a step further. It is essentially
categorical data for which there is a meaningful order. So it
increases or decreases in order. For example : If you ranked
all employees in an organization on the basis of their salary,
the rank you give each person would represent an order;
however, the actual amount of salary that each person (and
thus ranked position) differs by will not necessarily be equal.
Role level: 1) graduate, 2) consultant, 3) senior consultant, 4)

managing consultant, 5) vice principal
Job satisfaction: 1) very dissatisfied, 2) dissatisfied, 3) neither

satisfied nor dissatisfied, 4) satisfied, 5) very satisfied
Continuous variable types
• Continuous variables can take on any

numerical value on a measurement scale.
There are two key types of continuous
variable: interval variables and ratio variables.

Interval variables
• Interval data is measured on a numeric scale. The numbers
are continuous – so as well as knowing the order, interval data
records the exact value.
Daily temperature (Celsius): some examples might be 21.4, 37.8.

16.0, 8.5, etc.
Start date: some examples might be 6/6/1994, 28/9/2013,
4/5/2015, etc.
• Whilst we can measure the difference between two dates, we

cannot add or multiply this data.
• Interval data lacks a zero starting point, which makes
comparisons of magnitude not possible.
Ratio variables
• Ratio data is similar to interval data in that it is a continuous
numerical value but, unlike interval data, there is a defined
zero point.
Annual salary (£): Some examples might be 16000, 120000,

564482, 34909, 86269, etc.
Height (cm): 170, 182, 163, 134.5, etc.
Weight (kg): 67, 48, 120.3, 98.4, etc.
Job tenure (years): 0.25, 0.5, 1, 23, 20, etc.

Categorical variable types
Data Type Description Examples
Binary Only two categories Gender, someone who has left versus someone still
employed
Nominal Three or more categories Division or function (eg HR, Risk, Finance, Sales, etc);
country (eg Canada, Australia, India)
Ordinal Categories have a Customer satisfaction scale (very dissatisfied,

meaningful order dissatisfied, neutral, satisfied, very satisfied); ranked
sales position of employee or team (1st, 2nd, 3rd,
etc)

Continuous variable types
Data Type Description Examples
Interval Numerical value. No defined Temperature; year.

‘zero’ starting point.
Ratio Continuous numerical data Height; weight; salary; age; job

with a defined zero starting tenure; sales figures (€£$); % linked
point. Ratios of scores on a to an individual object such as
scale also make sense. individual, team, department (eg %
of the team who indicated that they
were ‘engaged’ or staff turnover %
in each team or division).

Data Analysis for Human Resources
• After data collection for HR research and HR decisions, we need to
organize and summarize data primarily with two statistical techniques,
i.e., measurement of central tendency and measurement of dispersion.
• Central tendency is measured calculating mean, mode and median.
• Mean is the average, mode is the value that occurs most and the
median is the mid value, midpoint or the 50th percentile.
• Dispersion or variability is measured using range, quartile deviation and
standard deviation.
• Range measures the difference between the highest and the lowest
score in a data set.
• Quartile deviation is the difference between the upper quartile and the
lower quartile in a data set. For example, if the upper quartile of a data
set is in 90th percentile, it means there are 90 percent scores below that
point. It can also be interpreted as: 90 percentile is in the top 10 percent
bracket.
• Standard deviation is the square root of the variance and it is the
distance of each score from the mean.
• Measurement of relative position indicates the performance score of
an employee in relation to others. This helps in understanding how well
the employee has performed compared to others.
• Two most frequently used measures of relative positions are percentile
ranks and standard scores.
• A percentile rank indicates the percentage of scores that fall at or
below a given score. If a score of 65 corresponds to the 80th
percentile, it means that 80 percent of the scores in the distribution are
lower than 65.
• A standard score is a derived score that expresses how far a given raw
score is from some reference point, typically the mean, in terms of
standard deviation units.
• The most commonly reported and used standard scores are z scores, t
scores. The z-score expresses how far a score is from the mean in
terms of standard deviation units. The t-score is expression of z-score
in different form.

Analysing Data: Histograms
• Frequency Distributions (aka Histograms)
– A graph plotting values of observations on the
horizontal axis, with a bar showing how many
times each value occurred in the data set.
• The ‘Normal’ Distribution
– Bell shaped
– Symmetrical around the centre

The Normal Distribution
The curve shows the idealized shape.

Properties of Frequency Distributions
• Skew
– The symmetry of the distribution.
– Positive skew (scores bunched at low values with the tail
pointing to high values).
– Negative skew (scores bunched at high values with the
tail pointing to low values).
• Kurtosis
– The ‘heaviness’ of the tails.
– Leptokurtic = heavy tails.
– Platykurtic = light tails.
Skew

Kurtosis

Central tendency: The Mode
• Mode
– The most frequent score
• Bimodal
– Having two modes
• Multimodal
– Having several modes

Bimodal and Multimodal Distributions

Central Tendency: The Median
• The Median is the middle score when scores
are ordered:
57 40 103 234 93 53 116 98 108 121 22 Data
22 40 53 57 93 98 103 108 116 121 234 Ordered Data
Median

Central Tendency: The Mean
• Mean
– The sum of scores divided by the number of
scores.
– Number of friends of 11 Facebook users.

The Dispersion: Range
• The Range
– The smallest score subtracted from the largest
– For our Facebook friends data the highest score is
234 and the lowest is 22; therefore the range is:
234 −22 = 212

The Dispersion: The Interquartile range
• Quartiles
– The three values that split the sorted data into four equal parts.
– Second Quartile = median.
– Lower quartile = median of lower half of the data
– Upper quartile = median of upper half of the data

Deviance
• We can calculate the spread of scores by
looking at how different each score is from the
center of a distribution e.g. the mean:

Standard Deviation
• The variance gives us a measure in units squared.
– In our Facebook example we would have to say that
the average error in out data was 3224.6 friends
squared.
• This problem is solved by taking the square root of the
variance, which is known as the standard deviation:

Populations and Samples
• Population
– The collection of units (be they people, plankton,
plants, cities, suicidal authors, etc.) to which we
want to generalize a set of findings or a statistical
model.
• Sample
– A smaller (but hopefully representative) collection
of units from a population used to determine
truths about that population

Statistical significance
• The basic idea behind statistical significance is linked to
hypotheses testing.
• This simply involves us developing a ‘research hypothesis’ and

what is called a ‘null hypothesis’.
• The test statistic tells us whether a particular pattern of

results can be relied upon and if it is robust enough for us to
make various assumptions about what is going on in our
population ; or whether it is more likely that we came by our
particular set of findings by chance alone.

Research hypothesis: attendance at a customer-service training
course by sales employees will increase customer satisfaction
survey scores.
Null hypothesis: attendance at a customer-service training

course by sales employees will have no impact on customer
satisfaction survey scores.
• If the test did indicate ‘statistical significance’, we would have grounds to

reject the ‘null hypothesis’ that the customer-service training course has
no impact on customer satisfaction survey scores.

• The criterion used universally for testing significance is 95 per cent
confidence.
• So if we can be at least 95 per cent confident (from statistical tests) that the
pattern of data found would be unlikely to occur by chance, we would say
that the finding is ‘statistically significant’.
• If we want to be 95 per cent certain about our findings then another way of
putting this is that we want to be able to say that there is only a 5 per cent
likelihood that our findings might be due to chance alone.
• The term used to indicate the level of significance in statistical tests is the
probability level. A 5 per cent likelihood is presented as a 0.05 probability.
This is called a p-value and this is what our statistical tests present: the
probability that our findings could be due to chance alone. So with the
statistical tests we want our p-values to be 0.05 or lower (commonly
indicated as p<0.05). Human Resource Analytics (18B12HS412) 85
• Sometimes our statistical tests indicate that we can be more
than 95 per cent confident that our findings are meaningful
and not due to chance alone. For example, our statistical test
may indicate that we can be 99 per cent confident that our
results would not occur by chance and thus the p-value
presented here will be p=0.01. If the p-value of the result is
smaller than 0.01 then we indicate p<0.01.

Statistical tests by data type
Dependent Independent Analysis Test Nature of the Test; what type of
Variable (DV) Variables (IV) research question is it trying to
explore?
Categorical Categorical and Chi-square analysis – Chi-square analysis explores

DV only one IV applied to a crosstabs of whether the frequencies and
(Binary, frequencies. proportions found across
Nominal, or particular categories (eg male
*Ordinal) versus female) might vary or be
linked to another category (eg
part-time versus full-time).


explore?
Categorical Continuous or Logistic regression analysis. Logistic regression analysis

DV together attempts to predict the
(Binary, continuous and likelihood of something in
Nominal, or categorical particular occurring (eg
*Ordinal) whether someone leaves the
organization versus whether
they stay) on the basis of
possible characteristics across
a range of independent
variables (eg performance
criteria, gender, salary
increases, promotion, etc).

explore?
Continuous Categorical- Independent samples t-test Independent samples t-
DV Binary IV test explores whether two groups
(Interval or Group A versus of data differ on some dimension.
Ratio – and Group B Eg are males more stressed than
sometimes females? Are females more
*Ordinal) engaged than males?
2-Category- Paired samples t-test Paired samples t-

Time Based test explores whether two
Binary IV – separate times/instances of
Group A at data collection of the same
Time 1 versus group differ or are the same
Group A at across time – has there
Time 2 been a change in some
data? Are the leadership
team performing better
than they were a year ago?

explore?
Continuous Categorical IV One-way independent ANOVA One-way independent
DV with three or ANOVA explores whether three
(Interval or more groups; or more groups of data differ on
Ratio – and Group A versus some continuous dimension, eg
sometimes Group B versus are the HR function’s employees
*Ordinal) Groups C etc… more engaged than those from
finance or marketing?
Category-Time One-way repeated One-way repeated

Based IV with measures ANOVA measures ANOVA explores
three or more whether three or more
time points separate times/instances of
(eg Group A at data collection differ or are
Time 1 versus the same across time – has
Group A at there been a change in
Time 2 versus some metric (eg team
Group A at engagement scores) across
Time 3) time?

explore?
Continuous Continuous Pearson’s correlation, or Pearson’s correlation, or Simple
DV (assumed) DV Simple linear regression linear regression – if you want to
(Interval or and only one look at the degree to which two
Ratio – and other continuous continuous variables are related in
sometimes IV some way – with correlations you
*Ordinal) may just be interested in the ‘co’-
relation between two variables;
with simple regression you might
be making some assumptions
about which variable could be
‘caused’ or ‘predicted’ by the
other.†

explore?
Continuous Time-based Repeated measures Repeated measures
DV categorical IV ANOVA (generalized linear ANOVA (generalized linear
(Interval or
Ratio – and with two or model). The DV metric will model) explores whether
sometimes more time have been collected on two or more separate
*Ordinal) points – Time more than two occasions times/instances of data
1 versus Time with the same objects (eg collection differ or are the
2 (etc) employees or teams) but same across time – has
in the sample can be broken there been a change in
combination up into one or more other some metric across time?
with one or category (eg male versus This approach also enables
more female; received training you to explore whether
categorical IVS versus not; HR versus particular comparator
(eg gender) marketing versus finance groups vary in the degree to
etc). which each differs across
time (eg male versus female;
received training versus not;
HR versus marketing versus
finance etc).

Dependent Independent Analysis Test Nature of the Test; what type of research
Variable (DV) Variables (IV) question is it trying to explore?
Continuous Two or more Multiple regression Multiple regression analysis –
DV continuous (linear) here we are trying to predict and
(Interval or IVs account for variation in a
Ratio – and continuous dependent variable (eg
sometimes sales €£$ figures for each
*Ordinal) individual – importantly we do this
by exploring the degree to which
this variation can be ‘accounted
for’ by the degree to which two or
more other (possible) causal
factors/IVS share variance with
this DV. For example – might the
variation in $ sales that an
employee achieves in a week be
linked to the number of hours
worked in a week and the age of
the employee?

Dependent Independent Analysis Test Nature of the Test; what type of research
Variable (DV) Variables (IV) question is it trying to explore?
Continuous Both Multiple regression Multiple regression analysis: as
DV categorical and (linear) immediately above – with the
continuous IVs
(Interval or difference that some of the IVs
Ratio – and included represent memberships
sometimes of a particular category/group (eg
*Ordinal) male versus female). If the
number of categories in a
particular IV is above two then
dummy variables have to be
created.

REFERENCES
Page, Limited, 2019
2017

Review Questions
1. A null hypothesis:
States that the experimental treatment will have an effect.
Is rarely used in experiments.
Predicts that an experimental treatment will have no effect on a
dependent variable of interest.
None of the above.

2. The aim of experimental research is to:
Be a phenomenon
Cause a phenomenon
Investigate what caused a phenomenon
Prevent a phenomenon

3. Ordinal level data are characterized by:
Data that can be meaningfully arranged by order of magnitude
Equal intervals between each adjacent score
A fixed zero
None of the above
4. In a typical research problem the is expected to influence the .
Predictor variable; primary variable
Independent variable; dependent variable
Dependent variable; independent variable
Criterion; hypothesis

5. If one is studying the impact of variable pay component on job satisfaction,
then variable pay component is
Moderating variable
Independent variable
Intervening variable
Dependent variable
Unknown variable
Extraneous variable

6. The salary of Ram is twice as much as the salary of Shyam – this is an
example of:
Nominal scale measurement
Ordinal scale measurement
Interval scale measurement
Ratio scale measurement

7. Which of the following scales possess an absolute zero?
Nominal scale
Ordinal scale
Interval scale
Ratio scale
None of the above

8. In which of the following scales the objects are arranged according to their
magnitude in an ordered relationship?
Nominal scale
Ordinal scale
Interval scale
Ratio scale

9. In which of the following scales can all possible statistical techniques be
applied?
Nominal
Ordinal
Ratio
Interval

10.The numbers assigned to the members of Team India is an example of
Nominal
Ordinal
Ratio
Interval

11. How would you conceptualize the measurement of blood pressure from a
group of individuals on multiple occasions to see if the results were sufficiently
similar?
Reliability
Validity
Similarity
Predictability

12. ________explores whether the frequencies and proportions found across
particular categories might vary or be linked to another category
Independent sample t-test
Multiple Linear Regression
Chi-Square Analysis
One-way Independent ANOVA

13. ________________analysis attempts to predict the likelihood of something
in particular occurring on the basis of possible characteristics across a range of
independent variables
Logistic Regression
Pearson correlation
Chi-Square Analysis

14. __________________ explores whether two separate times/instances of
data collection of the same group differ or are the same across time
Independent sample t-test
One-way Independent ANOVA
Paired sample t-test

15. ______________________ explores whether three or more separate
times/instances of data collection differ or are the same across time
One-way repeated measures ANOVA
Repeated measures ANOVA
One-way repeated measures ANOVA
None

DATA ANALYSIS TECHNIQUES

Chi-square (χ2) test
Observed gender and promotion values
2015 Promotions
Gender Promoted Not Promoted Total
Male 127 6784 6911
Female 49 8378 8427
Total 176 15162 15338
By combining gender and promotions in this way, we end up with

four discrete categories:
•male promoted;
•male not promoted;
•female promoted;
•female not promoted.

Chi-square
X2 = ∑ ((observed value – expected value)2/expected value)
•Male promoted:
Observed value = 127
Expected value that we would probably have if there was no link
between gender and promotions
= (total # men) × (total # promoted) / total people
= 6,911 × 176/15,338
= 79.3
Male not promoted:

Observed value = 6,784
Expected value
= (total # men) x (total # not promoted) / total people
= 6,911 x 15,162 / 15,338
= 6,831.7
Chi-square
•Female promoted:
Observed value = 49
Expected value
= (total # women) x (total # promoted) / total people
= 8,427 x 176 / 15,338
= 96.7
Female not promoted:

Observed value = 8,378
Expected value
= (total # women) x (total # not promoted) / total people
= 8,427 x 15,162 / 15,338
= 8,330.3
Chi-square
χ2 = ∑((observed value – expected value)2 / expected value)
= (observed male promoted – expected male promoted)2/expected male

promoted + (observed male not promoted – expected male not
promoted)2/expected male not promoted + (observed female promoted –
expected female promoted)2/expected female promoted + (observed
female not promoted – expected female not promoted)2/expected female
not promoted
= (127 – 79)2/79 + (6,784 – 6,832)2/6,832 + (49 – 97)2/97 + (8,378 –

8,330)2/8,330
= 2,304/79 + 2,304/6,832 + 2,304/97 + 2,304/8,330
= 29.16 + 0.34 + 23.75 + 0.28
= 53.53
Chi-square
This chi-squared (χ2) statistic is then checked against a χ2 distribution with
known properties in a table (online or in a stats book) along with the
degrees of freedom that, in this case, is calculated as (r–1)(c–1) where r is
the number of rows and c is the number of columns. So (2–1)(2–1) = 1 x 1
= 1.
When you look up what the critical values of a chi-square test for a 2 x 2
table (1 degree of freedom), these amount to the following:
p < 0.05 = 3.84
p < 0.01 = 6.63
p < 0.001 = 10.83
Our value of 53.53 is far greater than the chi-square value required for us
to be 99.9 per cent confident that our findings are not due to chance. We
can confidently say that as a proportion of males and females in our
organization, males get promoted significantly more frequently than
females. Human Resource Analytics (18B12HS412) 106
Ch-Square in SPSS
This will bring up another

box where you can select
the rows and columns of
your cross-tabulation (or
pivot) table. You will also
need to select the chi-
square statistic in the
‘Statistics’ button option
and select how you would
like the output table to
Human Resource Analytics (18B12HS412) look in the ‘Cells’ button
107
option.
Chi-Square Analysis
Observed Values
Male 127 6784 6911
Female 49 8378 8427
Total 176 15162 15338

Observed Values
Male 127 6784 6911
Female 49 8378 8427
Total 176 15162 15338
(row total * column total)

E = --------------------------------------
grand total
Observed Values
Male =176 * 6911/15338 =15162*6911/15338 6911

=79.3
=6,831.7
Female =176*8427/15338 =15162*8427/15338 8427

=96.70 =8330.30
Total 176 15162 15338

Chi-Square Analysis
O E (O-E) (O-E)^2 (O-E)^2/E
127 79.3 47.7 2275.29 28.69
6784 6831.7 -47.7 2275.29 0.33
49 96.7 -47.7 2275.29 23.53
8378 8330.3 47.7 2275.29 0.27
52.83
Degree of Freedom = (r-1) (c-1)= (1*1)=1

Chi-Square Calculated > Table Value
52.83 > 3.84
Interpretation: There is a significant

relationship between two categorical
variable

Question
support no support row totals:
conform: 18 40 58
not conform: 32 10 42
column 50 50 100
totals:
Calculate Chi-Square Value and interpret your results

Expected
Values
support no support row totals:
conform: 29 29 58
not conform: 21 21 42
column totals: 50 50 100
O E (O-E) (O-E)^2 (O-E)^2/E

18 29 -11 121 4.17
40 29 11 121 4.17
32 21 11 121 5.76
10 21 -11 121 5.76
19.87

t-Test
t (calculated) > t (table value)

• A research study was conducted to examine the differences between older and
younger adults on perceived life satisfaction. A pilot study was conducted to
examine this hypothesis. Ten older adults (over the age of 70) and ten younger
adults (between 20 and 30) were give a life satisfaction test (known to have high
reliability and validity). Scores on the measure range from 0 to 60 with high
scores indicative of high life satisfaction; low scores indicative of low life
satisfaction. The data are presented below. Compute the appropriate t-test.
Older Adults Younger Adults

45 34
38 22
52 15
48 27
25 37
39 41
51 24
46 19
55 26
46 36

• What is your computed answer (t (calculated))?
• What would be the null hypothesis in this study?
• What would be the alternate hypothesis?
• What is your tcrit?
• Is there a significant difference between the two groups?
• Interpret your answer.

Older Adults Younger Adults
45 34
38 22
52 15
48 27
25 37
39 41
51 24
46 19
55 26
46 36
Mean = 44.5 Mean = 28.1
S D= 8.68 SD = 8.54
S2 = 75.38 S2 = 72.98
t= (44.5-28.2)- (0)/ (75.38/10 + 72.98/10)^ 0.5 =4.25

• What is your computed answer (t (calculated))? 4.25
• What would be the null hypothesis in this study?
There is no significant difference between younger and older
adults on life satisfaction scores
• What would be the alternate hypothesis?
Life satisfaction scores of older and younger adults are different
• What is your tcrit?
t (table value)=2.101
• Is there a significant difference between the two groups?
Yes (t (calculated) > t (table value))

REFERENCES
Page, Limited, 2019
2017

Review Questions
1. In an attempt to compare the durability of two different materials (X and Y), 10 pieces
of type X and 14 pieces of type Y were used. From these samples we calculate the
statistics: ˉx=129.44x̄=129.44, ȳ=122.65ȳ=122.65, sx=9.15sx=9.15, sy=11.02sy=11.02. To
test for equal means, the appropriate t-distribution is Exactly one option must be correct
a.t24
b.t22
c.t10 - t14.
d.t23

3. A study was set up to look at whether there was a difference in the mean arterial blood
pressure between two groups of volunteers, after 6 weeks of following one of two
treatment programs. One group of volunteers were given an exercise regimen to follow
for the 6 weeks and the other group were given the same exercise regimen with the
addition of an experimental tablet.
Which type of t-test should be used in this situation?
a. One sample t-test
b. Independent samples t-test
c. Paired samples t-test
d. None of the t-tests would be suitable
4. A researcher asked 933 people what their favourite type of TV programme was: news,
documentary, soap or sports. They could only choose one answer. As such, the researcher
had the number of people who chose each category of programme. How should she
analyse these data?
e. t-test
f. One-way analysis of variance
g. Chi-square test
h. Regression

7. What is the purpose of a chi-square test?
a. To measure the degree of association between categorical variables.*
b. To distinguish the degree of separation between categorical variables.
c. To evaluate the degree of differentiation between continuous variables.
d. To quantify the degree of differentiation between categorical variables.
Calculate the totals for the table below.

8. Calculate the expected value for the athletes who were inspired.
a. 142*
b. 146
c. 150
d. 138

9. Calculate the chi-square value for the above example.
a. 2.876*
b. 2.600
c. 2.881
d. 0.586

10. Which of the following is the correct calculation for the degrees of freedom (df) when
using a 2 × 2 contingency table?
e. (n – 1) × (n – 1) = 1*
f. (n + 1) × (n + 1) = 9
g. (n – 1) + (n – 1) = 2
h. (n + 1) + (n + 1) = 6

11. Identify the critical value from a chi-square distribution table based on the df
calculated above and a significance level of .05.
a. 3.84*
b. 5.99
c. 7.81
d. 9.49

12. Based on the chi-square value calculated and the critical value identified above, what
can be deduced from the association between athletic status and whether the London
Olympic Games inspired them to become more physically active?
e. A significant difference exists between athletic status and the level of inspiration.
f. No significant difference exists between athletic status and the level of
inspiration.*
g. The relationship between athletic status and the level of inspiration is unclear.
h. A minor significant relationship exists between athletic status and the level of
inspiration.

ID Group Days
Analyse the data below related 1 Stomach 223
to days survival following 2 Stomach 224
3 Stomach 151
surgery from either stomach or 4 Stomach 300
lung cancer. 5 Stomach 246
6 Stomach 166
7 Stomach 100
8 Stomach 111
13. Run an independent t-test on 9 Stomach 155
the data. A Levene’s test result of 10 Stomach 247
11 Stomach 151
p = .006 was obtained. What can 12 Stomach 166
we infer from this number? 13 Stomach 99
a. The stomach and lung cancer 14 Bronchus 223
15 Bronchus 138
variance is heterogeneous.* 16 Bronchus 72
b. The stomach and lung cancer 17 Bronchus 245
variance is homogenous. 18 Bronchus 124
19 Bronchus 368
c. The distributions look fairly 20 Bronchus 112
similar. 21 Bronchus 555
d. The result is inconclusive. 22 Bronchus 412
23 Bronchus 111
24 Bronchus 1112
25 Bronchus 479
26 Bronchus 103
27 Bronchus 876
28 Bronchus 146
29 Bronchus 340
Human Resource Analytics30
(18B12HS412) Bronchus 396 128
14. What is the correct result for the independent t-test?
a. t(17.95) = –2.24, p = .038*
b. t(28) = –1.98, p = .058
c. t(17.95) = –2.24, p = .006
d. t(28) = –1.98, p = .006

15. If a significance result of .000 is presented, how should this be reported?
e. p = .000
f. p < .000
g. p < .001*
h. p > .001

DATA ANALYSIS TECHNIQUES


REGRESSION ANALYSIS
Linear regression is based on

1. Fitting a straight line to data
2. Explaining the change in one variable through changes in other variables.
Regression Techniques :
Y = a + bX
Independent
variable
Slope
Dependent Vertical i.e., marginal change
variable intercept
Testing Regression Estimates :
The strength of relationship between the dependent variable and
the independent variables can be measured in two ways :
1. The Coefficient of Determination (R2) :

- used to measure how well the overall equation explains
changes in the dependent variable.
2. The t-statistic :
- used to test the strength of the relationship between an
independent variable and the dependent variable.

Coefficient of Determination
R 
2 Explained Variation

 (Yt  Y )
ˆ 2
TotalVariation  (Yt  Y ) 2

t Statistic
The t-test is used to determine if there is significant relationship

between the dependent variable and the independent variable.
bˆ 3.53
t   6.79
sbˆ 0.52

If the absolute value of t > tabular value
oder
The value of standard error < tabular value,
there exists a statistically significant relationship

between the two parameters

Multiple Linear Regression-HR
Analytics
• Dependent Variable:
BAME :Percentage of the unit made up of

Black, Asian or Minority Ethnic employees

Independent Variables:
GroupSize Number of employees in the unit
PercentMale Percentage of the unit made up of males
NumberFeMaleTeamLeads Number of team leads in the unit who are female
LondonorNot Geographical location recoded to 1 = Central London or Greater

London and 2 = Rest of UK
Function Function 1 = Sales staff (customer-facing people) or 2 =

Professional Service (non-customer-facing people)

Analyze-------Regression-----------Linear

REFERENCES
Page, Limited, 2019
2017

Review Questions
1. Anne is investigating the influence of factors such as room temperature, humidity
and room lighting on how many mistakes an individual makes on a computer-based
reaction time game. Which of the following statistical tests best suits her study
design?
a. Linear regression
b. Chi-square test
c. Pearson correlation
d. Spearman rank order correlation

3.Which of following statement is true?
a. 0 < R2 < 1
b. R2 > 1
c. R2 = 1
d. -1 < R2 < +1

4.When will the value of R2 equal one?
e. When the difference between actual values of dependent variable and the
error term is zero for all observations
f. When there is no measurement error in the dependent variable
g. When the estimated value of the error term for all observations is
zero
h. None of the above statements is true

5.Zero correlation coefficient between two variables could mean
i. The variables are non-linearly related to each other
j. There is a cause and effect relationship between variables
k. That there is error of measurement in variables
l. None of the above is true

6.The slope term of simple linear regression cannot be estimated if
a. All values of the independent variable are same
b. Variance of independent variable is maximized
c. The values of independent variable is the reciprocal of the value of
dependent variable
d. None of the above statement is correct

7. The multiple regression model cannot be estimated if
e. Simple correlation coefficient between dependent and any of the
independent variable is 1
f. If correlation coefficient between any two independent variables is zero
g. If correlation coefficient between any two independent variables is
h. Both (b) & ( c ) are true


9. Which of following measures the explanatory power of the regression model?
a. Correlation coefficient
b. Coefficient of determination
c. Regression coefficient
d. All of the above

10.For the regression model Y =b0 + b1 X1 + b2 X2 +U
e. X1 X2 are dependent variables.
f. Y is a predictor variable.
g. b0 is predictor variable.
h. None of the above is true.

11. If the estimated regression equation is
Y = 15 - 0.2 X, where Y = quantity demanded and X = price, then
i. When X =10, Y =12
j. When X =0, Y =15
k. When X =20, Y =11
Both ( b ) & ( c ) are correct

12. Which of the following value of the correlation coefficient indicates a weak
relationship?
a. 0.09
b. 0.54
c. -.76
d. -.89

13. The regression model is expressed as:
e. Y = a + b X
f. Y = a + b X + U
g. Ŷ = + X
h. Ŷ = a + b X + U

14. If correlation coefficient between X and Y is zero, it indicates that
i. X and Y are unrelated
j. X and Y have a relationship
k. X and Y do not have a linear relationship
l. None of the above

15. The significance of R2 is conducted using
a. t statistic
b. Z statistic
c. Chi-square statistic
d. F statistic

HRA-Chap 1 (2020)

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

HRA-Chap 1 (2020)

Uploaded by

Copyright:

Available Formats

HR ANALYTICS-CHAPTER 1

Human Resource Analytics

Human Resource Analytics

Human Resource Analytics

2. Human Resource information Understanding HR metrics and data, Data 8

Human Resource Analytics

Human Resource Analytics

Human Resource Analytics

Human Resource Analytics

• “HR analytics is a methodology for creating insights on how

Human Resource Analytics

Human Resource Analytics

There is a great deal of discussion on replicating the

Human Resource Analytics

Human Resource Analytics

2. Cost per hire (total cost of hiring/the number of new hires)

3. Early turnover (percentage of recruits leaving in the first year)

5. Revenue per employee (revenue/total number of employees)

7. Billable hours per employee

9. Cost of HR per employee (e.g. $ 600)

10. Ratio of HR professionals to employees (e.g. 1:60)

12. Turnover (number of leavers/total population in the organization)

13. Effectiveness of HR software

14. Absenteeism (absence percentage)

Human Resource Analytics

•Current churn rate

Human Resource Analytics

As qualified candidates become more difficult to find and retain,

Human Resource Analytics

Human Resource Analytics

Over time, HR analysts can also see

HR analytics is also playing an increased

The best employee development and talent management programs align

Human Resource Analytics

Human Resource Analytics

Google is a perfect case study in how people analytics can be used

Human Resource Analytics

Incentive programs to motivate employees to maximize productivity. Data

Human Resource Analytics

Executive compensation analysis needed to attract and retain top leaders.

Human Resource Analytics

•Better hiring practices

Human Resource Analytics

Human Resource Analytics

Human Resource Analytics

This is also the reason why we put SPSS on the

Human Resource Analytics

RStudio is an open source and enterprise-ready

Human Resource Analytics

Excel is where most of us started. It’s no

The good thing about Excel is that it’s very

For example, if you wanted to check how

Human Resource Analytics

Bhattacharyya, HR Analytics: Understanding Theories and Applications, Sage, 2017

Human Resource Analytics

3. _________ is an enterprise analytics 4. _________ is an open-source

Human Resource Analytics

Human Resource Analytics

15. The most popular web analytics tool is

DATA & ANALYSIS STRATEGIES

Human Resource Analytics (18B12HS412) 44

• HR research can be either quantitative or qualitative.

• Quantitative data are gathered through; experimental/clinical trials,

3. _ is an enterprise analytics 4. _ is an open-source