Unit 1.12 - HR Research and Analytics Flashcards
How can the HR department contribute to better management decisions using research?
The HR department can contribute to better management decisions through an effective research program, using both quantitative and qualitative research designs to answer important questions.
Collection of HR Data:
* Employee demographics
* Job satisfaction
* Employee skills
* Compensation programs
* Training needs
* Safety records
What are 3 major characteristics of the scientific method (that separates it literature, art, etc. analysis)?
The scientific method is a systematic approach of developing and testing new ideas.
- Hypothesis - a proposed explanation between two variables based on limited evidence
- Constructs - a subjective idea, concept, object or event not based
on evidence - Observations - Gathered information from constructs that may support the hypotheses
What are 4 of the most popular research methods in experimental design?
- Case study - observational method; poor research design because of inadequate experimental controls
- Field survey - gathering info from employees who are working in their natural, everyday environment; correlational study (ex. age and work accidents)
- Field experiment - gathering info from employees in their natural environment + manipulating one variable (the independent variable) and analyzing its effects on other variables (the dependent variables); allows experimenters to infer causality (ex. testing whether financial incentives cause employees to work more safely)
- Lab experiment - individuals are brought into a research lab and participate in a standardized test; most rigorous research design, but it doesn’t mean it’s the best, given it’s artificial environment and difficulty of use
What are the 7 most frequently used research designs?
X = experimental treatment (independent variable)
O = observation (dependent variable)
R = randomization (each R is a randomly assigned group of people)
- Post-Test Only (Case Study) -
Example: conduct training program, observe satisfaction with training
X O - Pre-Test Post-Test Comparison -
Example: observe satisfaction with training, conduct training program, observe satisfaction again
O X O - Pre-Test Post-Test Control Group -
Involves two groups: experimental and control with participants randomly assigned to each group
R O X O
R O O - Post-Test Only Control Group -
Like #3 but with no pre-test for either group
R X O
R O - Solomon Four Group Design -
Most highly recommended research design for evaluating training programs; 4 randomly assigned groups; 2 groups get the training
R O X O
R O O
R X O
R O - Time Series Design -
Consists of periodically measuring something and introducing experimental change during the series of measurements
O1 O2 O3 O4 X O5 O6 O7 O8 - Separate Sample Pre-Test Post-Test Control -
Utilizes two existing groups that may not be equivalent; only one group receives the experimental change
Group 1:
R O X
R X O
Group 2:
R O
R O
The first two aren’t great experimental designs to use
The last two are called quasi-experimental designs because they occur in a natural setting where you can introduce an experimental change and collect data, but don’t have complete control over the environmental setting
What are the 4 kinds of data available in HR research?
- Archive Data - info contained in the HR files; sometimes referred to as unobtrusive measures because assessing and obtaining it doesn’t affect employees’ behavior (ex. past attitude surveys, insurance records, disciplinary records)
- Observations - observing behavior; a disadvantage is that it’s obtrusive; can cause the Hawthorne Effect, which is a change in employee behavior because they’re being observed (ex. working faster)
- Interviews - can provide information about a broad range of topics and reveal personal opinion, but they are time consuming and only a limited number of employees can be interviewed; focus groups are useful for this
- Questionnaires - can be administered to a large sample of employees, but they are time consuming and only answer questions included on the questionnaire
Define reliability and validity
All data collection methods must be both reliable and valid.
Reliability - the consistency or repeatability of the measure
Validity - concerned with whether the research instrument actually measures what it is supposed to be measuring
What is a frequency distribution and normal distribution?
Frequency distribution - reports the results of a group of numbers (ex. bar chart)
A bell-shaped pattern of responses, called a normal distribution, occurs when almost any human characteristic is measured over a large population of people
What are the measures of central tendency?
In descriptive statistics
Mean (aka average)
Median
Mode
What is the degree of dispersion?
In descriptive statistics
Indicated by:
Variance - indicates whether a group of numbers are all close to the mean or whether they’re widely scattered above and below the mean
Standard deviation - square root of the variance
Range - the distance from the lowest number to the highest number
What are descriptive statistics?
Used to summarize data in meaningful ways so the results can be used to make decisions
Measures of central tendency:
Mean
Median
Mode
Degree of dispersion:
Variance
Standard deviation
Range
What are inferential statistics?
Used to make inferences about a larger population of people when
data have been obtained from only a small sample.
Two of the best procedures for making accurate inferences:
Simple random sample - placing all employees in the sample population and drawing the sample at random
Stratified random sample - categorizing employees into specified groups according to relevant characteristics, such as team or job level
Explain correlation and regression
A correlation coefficient is a number that shows the relationship between two variables (always a number between -1.0 and +1.0; a correlation of 0 means non relationship between variables); does not infer causality
A regression analysis shows the relationship between one variable and several other variables.
What is qualitative research (rather than quantitative)?
Qualitative research focuses on observations and interviews with specific individuals in unique settings (ex. ethnographic research; a lot comes from the legends and myths members of the group tell)
List the graphs and charts frequently used by HR managers
Histogram -
- ex. can be used to show test scores of participants in a training program, with each bar representing a range of test scores
Column chart -
ex. represent the average longevity on the job of employees in each department (distinct bars, unlike histograms(
Bar chart -
Like column charts, but have categories listed along the vertical axis and numbers along the horizontal axis
Pie chart -
ex. percentage of races of employees in a department (must equal 100%)
Line graph -
Frequently used to show trends or progress over a period of time (time is showed on horizontal axis, values on vertical axis)
Dual-axis chart -
Used to show the relationship between two sets of data
ex. it could be bar chart and line graph on top of each other, such as one representing % of safety trained, and one of them accidents per month
Scatter plot -
Shows trends in the distribution of data; displays relationship between two variables (ex. test scores and job performance)
Venn diagram -
Two or more overlapping circles that show the relationship between sets; shows what the sets have in common
Gantt chart -
Displays all the tasks involved in a project, the relative order of the tasks, and the amount of time each task is expected to take
