Chapter 5

Question 1

HR demand refers to:

Answer 1

> HR demand refers to the firm’s future need for human capital, and to the types of jobs and number of positions that must be filled for the firm to implement its strategy.

Question 2

The demand for human capital is determined by:

Answer 2

> by the strategic and operational requirements of the firm or business unit.

> This means that understanding the demand for talent begins with the firm’s strategy, and flows from the value-generating activities of the firm.

Question 3

Because of the variety in the complexity and levels of uncertainty in forecasting the demand for labour, multiple forecasting methods exist. These methods can be divided into two main categories:

Answer 3

> quantitative methods and qualitative methods

Question 4

The two main factors that determine whether a quantitative model or qualitative model is a better choice are the:

Answer 4

> degree of uncertainty involved in the demand forecast, and the volume and complexity of the data that are available to assist in creating the demand forecast.

Question 5

What is fluid work?

Answer 5

> The job is being broken apart into more loosely organized groups of tasks, and many tasks are being taken out of the job altogether and performed by specialized workers or outsourced to the gig economy. This is referred to by some scholars as “fluid work,” and it is evidenced in the workplace by the increased use of gig workers, contractors and other forms of outsourcing of tasks, consultants for specialized tasks, part-time workers, and job-sharing practices

Question 6

What is the first step in assessing the demand for labour?

Answer 6

> This is the first step in assessing the demand for labour, as it gives the organization a valid and reliable understanding of what work needs to be done.

Question 7

Some quantitative models are based on something that is known. Describe this:

Answer 7

> Some quantitative models are based on what is known about existing relationships between a level of consumer demand or production, where a forecast already exists, and human capital demand.

Question 8

While no forecast is ever perfectly accurate, it may be possible to improve the accuracy of forecasting by including more factors that contribute to changes in demand. Provide some examples:

Answer 8

> customer demand for some products may be highly seasonal, and so production and consequently the demand for labour will change seasonally.

> Rather than basing demand forecasts on total sales per year, labour demand forecasts can be produced that follow product sales more closely when the seasonality of sales is included.

Question 9

Can consistency be increased in forecasts? Is their a price to pay for increasing it?

Answer 9

> by observing how the demand for labour changes, theorizing what might influence changes in demand, and collecting data based on theory and observations, it may be possible to increase the consistency of forecasts. H

> However, this increase in consistency comes at the price of collecting data over longer periods of time and collecting data from a wider variety of sources.

Question 10

quantitative models are better when forecasting what?

Answer 10

> In general, quantitative models are better when forecasting demand in stable markets when there is a high degree of certainty in the relationship between the demand for labour and the indicators of that demand.

Question 11

Provide a summary of the trend/ration analysis?

Answer 11

> This uses historical changes to predict future human capital needs.

Question 12

Provide a summary of time series model:

Answer 12

> These models use past data to predict future demand.

> Time series models are especially useful for capturing seasonality in data (such as a December rush on sales and a summer slump).

Question 13

Provide a summary of correlation analysis:

Answer 13

> This method describes the strength of a relationship between two variables (for example, job satisfaction and job commitment, or sales and number of employees).

Question 14

Provide a summary of regression analysis:

Answer 14

> This shows a linear relationship or trend between one or more predictor variables and an outcome variable.

> A regression analysis allows the user to predict the outcome based on known values of the predictor variables.

Question 15

Provide a summary of the Structural Equation Modelling (SEM):

Answer 15

> This modelling shows the relationships between multiple predictor and outcome variables.

> SEM is useful for examining more complex models with more variables than regression models.

Question 16

Provide an overview of Machine Learning and Artificial Intelligence (AI)

Answer 16

> This finds patterns in large amounts of data (for example, Amazon customer product reviews).

> Machine learning can make use of many more variables than regression or SEM, and can be used to predict a particular outcome, or identify categories among large amounts of data.

> Categorization-Based
Prediction-Based
Neural Networks
Random Forests

Question 17

What is trend analysis?

Answer 17

> Trend analysis is a general term for any type of quantitative approach that attempts to forecast future human capital needs by extrapolating from historical changes in one or more organizational indices. A basic form of trend analysis could be plotting previous levels of employment to determine future needs

Question 18

What is ratio analysis?

Answer 18

> Ratio analysis involves examining the relationship between an operational index and the demand for labour (as reflected by the number of employees in the workforce) and is a relatively straightforward quantitative demand forecasting technique commonly used by many organizations

Question 19

Although sales level is probably the most common index used by organizations, other operational indices include (Trend/Ratio):

Answer 19

(1) the number of units produced,

(2) the number of clients serviced, and

(3) the production (i.e., direct labour) hours.

Question 20

some organizations use ratio analysis to ascertain demand requirements for what specifically?

Answer 20

(1) direct labour and

(2) indirect labour (e.g., HR staff).

Question 21

What are the five steps to conducting an effective ratio analysis.

Answer 21

1) Step 1. Select the Appropriate Business/Operational Index. The HR forecaster must select a readily available operational index, such as sales level, that is (1) known to have a direct influence on the organizational demand for labour and (2) subjected to future forecasting as a result of the normal business planning process.

2) Step 2. Track the Operational Index Over Time. Once the index has been selected, it is necessary to go back in time for at least the four or five most recent years, but preferably for a decade or more, to record the quantitative or numerical levels of the index over time.

3) Step 3. Track the Workforce Size Over Time. Record the historical figures of the total number of employees, or, alternatively, the amount of direct and indirect labour for exactly the same period used for the operational index in Step 2.

Step 4. Calculate the Average Ratio of the Operational Index to the Workforce Size. Obtain the employee requirement ratio by dividing the level of sales for each year of historical data by the number of employees required to produce that year’s level of sales. This ratio is calculated for each year over the period of analysis so that an average ratio describing the relationship between the two variables over time can be determined.

Step 5. Calculate the Forecasted Demand for Labour. Divide the annual forecast for the operational index by the average employee requirement ratio for each future year to arrive at forecasted annual demand for labour. For example, obtain future sales forecast figures for the next five years. For each of the years, divide the level of sales by the average employee requirement ratio to obtain the forecasted numerical demand for labour for each future year (Figure 5.4).

Question 22

What are time series models?

Answer 22

> Times series models use past data to predict future demand. They can range from very simple to highly complex.

Question 23

What are the two different averages that can be used in the time series model?

Answer 23

> simple moving average

> weighted moving average, in which all periods of actual demand data are used to estimate future demand, but greater weight is given to more recent demand data. A weighted moving average places more importance on recent demand data.

Question 24

What is correlation?

Answer 24

> Correlation is used to describe the relationship between two variables.

> where an increase in one variable associates with a decrease in the other variable, such as how higher job satisfaction associates with lower turnover. If there is no relationship between two variables (for example, if changes in job satisfaction are not expected to have any association with changes in cognitive ability) then the correlation is zero. Thus, correlation describes the strength of the association between two variables, and it can range from 1, which represents a perfect positive relationship, to negative 1, which represents a perfect negative relationship.

Question 25

Correlation provides information around :

Answer 25

> Correlation provides information around the direction and strength of the association between two variables.

Question 26

What does shared variance allow us to understand and what does it mean?

Answer 26

> the concept of shared variance allows us to understand how much of one variable can be predicted by the other variable. > Shared variance is simply the correlation coefficient squared; and since a correlation coefficient is a number between -1 and 1, the shared variance, or R2, is a smaller number than the correlation coefficient (unless the correlation coefficient is exactly -1 or 1).

Question 27

What is regression analysis?

Answer 27

> Regression analysis extends the idea of correlation in several important ways. 1) First, while correlation describes the strength of a relationship in values that range from -1 to 1, regression describes the strength of a relationship in the same values as the actual variables, which permits a more straightforward interpretation. 2) Second, regression can integrate the combined effects of multiple variables on another variable.

Question 28

Regression analysis can be thought of loosely as an extension of what?

Answer 28

> Regression analysis can be thought of loosely as an extension of trend analysis, except that rather than using a single operational index (which in regression would be called the independent or predictor variable), regression uses more data than the ratio of the operational index to employees, and regression can use multiple predictor variables to forecast the required number of employees.

Question 29

What is the purpose of regression analysis?

Answer 29

> The purpose of regression is to estimate a trend or relationship between one or more predictor variables and an outcome variable

Question 30

The best regression model is what?

Answer 30

> The best regression model is the simplest model that predicts the most variability in our criterion variable. Thus, the best model given these data is our first model, which uses production as a single predictor of employees.

Question 31

When developing a model, Meehan and Ahmed (1990, 297–307) suggest grouping possible predictor variables in three categories:

Answer 31

1) key variables, which are almost certain to play a role in predicting the employee requirement; 2) promising variables, which appear important and you suspect will probably relate to predicting the employee requirement; and 3) possible variables, which may not be necessary for the model.

Question 32

While regression models can be highly useful forecasting tools, like any powerful tool, statistical regression has its limitations:

Answer 32

> Unsuitable for nonlinear relationships. > Large amounts of historical data are required.

Question 33

Regression models help HR planners to:

Answer 33

> Regression models help HR planners to make use of large amounts of organizational data.

Question 34

What is Structural equation modelling (SEM)?

Answer 34

> Structural equation modelling (SEM) can be thought of as a process similar to regression, except that where regression deals with a single outcome variable at a time, there can be many outcome variables in a single SEM model.

Question 35

The primary drawback to SEM versus regression is:

Answer 35

> The primary drawback to SEM versus regression is that SEM typically requires more data observations than regression.

Question 36

One of the largest potential benefits of artificial intelligence and machine learning tools is the:

Answer 36

> ability to find patterns in large amounts of data that would otherwise not be detectable.

Question 37

How do AI methods sort data?

Answer 37

> These methods sort data into categories based on similarity.

Question 38

Categorization-based machine learning methods may be useful to:

Answer 38

> Categorization-based machine learning methods may be useful to identify patterns among employee data as well.

Question 39

Many prediction models are called what?

Answer 39

. any prediction models are called supervised models because the analysis is based on the prediction of a particular outcome.

Question 40

What are neural networks?

Answer 40

> Neural networks are named for their resemblance to the structure of the neural pathways in our brains. Neural networks are made up of many nodes, where a node is a point at which information arrives and is acted upon before being passed on as another piece of information for another node.

Question 41

There are many different types of supervised and unsupervised machine learning methods. The common aspects of machine learning models in general are that they learn by being trained on data where the predictors and the value of the outcome are already known, and then validated against a portion of the data that was withheld from the training stage. What does this ensure?

Answer 41

> This ensures that the algorithm is predicting with a high degree of accuracy and that the algorithm has not been over-trained on the data. This is partly why the validation stage is so important when developing a machine learning model.

Question 42

What are qualitative models preferable? Is it common?

Answer 42

> When there is a great degree of uncertainty, qualitative models may be preferable. > Qualitative models may also be useful when no formal planning exists, and no formal data collection occurs around a planning process. > Qualitative forecasting in the form of direct managerial input is the most commonly used method for determining workforce requirements

Question 43

Using experts to arrive at a numerical estimate of future labour demand is considered to be what kind of process?

Answer 43

> Using experts to arrive at a numerical estimate of future labour demand is considered to be a more formal qualitative process for determining future labour requirements, because it is a detailed process of stating assumptions, considering potential organizational and environmental changes, and deriving a rationale to support the numerical estimate.

Question 44

One of the issues with qualitative decision methods, as pointed out by Philip ­Tetlock

Answer 44

> that experts are very often wrong. > groups are often more correct.

Question 45

Qualitative forecasting methods such as the Delphi technique and nominal group technique (NGT) make use of what?

Answer 45

> make use of groups of experts in an effort to increase the validity and reliability of the forecast and of the information used to formulate the forecast. > . Furthermore, qualitative methods of forecasting such as these increase the reliability of the information by working through an iterative process where each expert has the opportunity to clearly explain his or her assumptions and decisions.

Question 46

What is scenario planning?

Answer 46

> Scenario planning is a method often used to develop organizational strategy. A primary strength of scenario planning as a strategy-setting tool is that it encourages participants to develop strongly shared mental models of future organizational states. Scenario planning is a method for imagining future possible conditions in which the organization might operate. As a technique, it requires participants to challenge existing assumptions and to generate vivid pictures of possible future states.

Question 47

What is the general process for scenario planning?

Answer 47

1) Propose the forecasting question about the future state of the firm or ­environment. For example, “How many households will own an electric car in 10 years?” 2) Generate a list of factors that are likely to influence the outcome in question. It is often useful to perform a SWOT analysis that takes into account factors such as the economy, the political landscape, society, and the impact of technology. 3) Sort the factors into naturally occurring groups and rank the groups according to their importance to the main question and the ability of the firm to control the factor. Factors that the firm has less ability to control should receive higher rankings. 4) Select the two groups of factors that are likely to have the strongest and most unpredictable impact on the question. Create four quadrants (refer to Figure 5.19) by stretching one group along a continuum from its extreme negative condition to its extreme positive condition on the x-axis, and stretching the other group along a continuum from its extreme low to high conditions on the y-axis. 5) Name and describe in story form each of the four worlds in the four resulting quadrants. 6) Suggest the skills, competencies, and other organizational requirements that would be necessary for the firm to be able to operate in each of these four worlds. 7) Generate a demand forecast necessary to fulfill the firm’s requirements in each of the four worlds.

Question 48

The benefit that scenario planning brings to demand forecasting is:

Answer 48

The benefit that scenario planning brings to demand forecasting is that while there is really no limit to the number of possible scenarios that the future could hold, this method allows planners to understand the most important assumptions that go into each demand forecast.

Question 49

What is the delphi technique?

Answer 49

> The Delphi technique, named after the Greek oracle at Delphi and developed by N. C. Dalkey and associates at the Rand Corporation in 1950, is another useful qualitative method for deriving detailed assumptions of long-run HR demand > This forecasting technique is “a carefully designed program of sequential, individual interrogations (usually conducted through questionnaires) interspersed with feedback on the opinions expressed by the other participants in previous rounds > Instead, a project coordinator canvasses them individually for their input and forecasts by means of a progressively more focused series of questionnaires.

Question 49

What are advantages to the delphi technique?

Answer 49

> The advantage of the Delphi technique is that it avoids many of the problems associated with face-to-face groups, namely reluctance on the part of individual experts to participate due to (1) shyness, (2) perceived lower status or authority, (3) perceived communication deficiencies, (4) issues of individual dominance and groupthink > Because the Delphi technique does not employ face-to-face meetings, it can serve as a great equalizer and can elicit valid feedback from all expert members. It is also advantageous that the Delphi technique can effectively use experts who are drawn from widely dispersed geographical areas

Question 50

What are disadvantages to the delphi technique?

Answer 50

> s. There are disadvantages associated with the Delphi technique, as there are with all forecasting techniques. In particular, because of the series of questionnaires administered to derive a forecast, the time and costs incurred when using the Delphi technique can be higher than those incurred when using alternative forecasting methods. > Another deficiency is that since the results cannot be validated statistically, the process is greatly dependent on the individual knowledge and commitment of each of the contributing experts > Furthermore, if the experts are drawn from one specific field, their common professional training might guide them along a single line of inquiry > if insufficient attention has been paid to developing criteria for the identification and selection of experts, the people selected to derive the demand forecasts may lack sufficient expertise or information to contribute meaningfully to the process

Question 51

What are the 6 steps to the delphi process?

Answer 51

1) Define and refine the issue or question 2) identify the experts, terms, and time horizon 3) Orient the experts 4) issue the first round questionnaire 5) Issue the First-Round Questionnaire. 6) Continue Issuing Questionnaires.

Question 52

What is the nominal group technique?

Answer 52

> Although the nominal group technique (NGT) is also a long-run qualitative demand forecasting method, it differs from the Delphi technique in several important respects. First, unlike in the Delphi technique, the group does, in fact, meet face to face and interact, but only after individual, written preparatory work has been done and all the demand estimates (idea generation) have been publicly tabled, or written on a flip chart, without discussion (Figure 5.21) (Fraser and Fraser 2000, 228–243; Van de Ven 1974). Second, each demand estimate is considered to be the property of the entire group and to be impersonal in nature, which minimizes the potential for dominance, personal attacks, and defensive behaviour in support of estimates presented in the group forum (Rohrbaugh 1981, 272–288). Finally, the expert forecast is determined by a secret vote of all group members on their choice of the tabled demand forecasts.

Question 53

Studies have shown that nominal group technique (NGT) is especially effective for:

Answer 53

> Studies have shown that nominal group technique (NGT) is especially effective for brainstorming sessions to ensure all participants have an equal voice in the sessions, and when a problem or issue stems from several widely diverse causes. Furthermore, studies have shown that NGT provides highly reliable and valid qualitative data that is ranked by importance and is superior to that derived from focus group sessions.

Question 54

What are HR budgets?

Answer 54

> HR budgets are quantitative, operational, or short-run demand estimates that contain the number and types of jobs or positions required by the organization as a whole and for each subunit, division, or department.

Question 55

What does the HR budget produce?

Answer 55

> The HR budget process produces a staffing table, which contains information related to a specific set of operational assumptions or levels of activity (e.g., maintain the current organization structure, increase the sales level by 5 percent over last year’s level)

Question 56

Are qualitative and quantitative methods exclusive?

Answer 56

> Quantitative and qualitative methods are not exclusive methods where the forecaster decides to use one method at the expense of the other. These two basic methods can both be used in order to test the reliability of the forecast. If both methods arrive at similar forecasts, then planners can be more certain of the results. However, if one method leads to dramatically different results from the other, then it would be reasonable to question the assumptions of both methods to try to understand where differences in assumptions lead to discrepancies in the forecasts.

Question 57

What are the data differences in qualitative and quantitative?

Answer 57

QUANTITATIVE: - Economic outlook - labour force growth - labour force projections - industry projections - occupational projections QUALITIATIVE: - economic outlook - demographics - political and regulatory environment - technological change - societal changeg

Question 58

What are the result differences in qualitative and quantitive?

Answer 58

Quantitative - Detailed, numeric estimates Qualitative - More specific information that can easily be used to produce HR system development plans

Question 59

What is simulation?

Answer 59

> Simulation is a powerful blend of quantitative and qualitative analysis. It is based on the creation of a set of assumptions around the variables or inputs that are expected to affect demand followed by a quantitative representation of those assumptions and their interactions. These assumptions start at the macro level (e.g., the economy, changes in technology, political or legislative outcomes, or demographics), and then incorporate organizational variables that could also influence demand such as the impact of technology or process changes in production. Simulation can be used to model both demand and supply, so models can also include supply-related variables such as the amount of time required for training someone to become competent at a job level, and seniority levels of employees in the job.

Question 60

The assumptions that are used to build a simulation model employ :

Answer 60

> The assumptions that are used to build a simulation model employ mathematical algorithms that reflect how variables are expected to react dynamically with the other variables in the model.

Question 61

What kind of data can non-simulation show?

Answer 61

> simulation can model these nonlinear dependencies, and so simulation has an advantage over regression and trend-based models for these types of applications.

Question 62

What are the steps to simulation?

Answer 62

1) Using qualitative methods, collect the relevant variables. This process can use interviews, focus groups, Delphi technique, or NGT to collect the information. 2) Describe how these variables interact together to by developing a process model to map the relationships between variables. Line managers who have direct experience with the manner and ways in which the demand for human capital changes over time and with sales are in an excellent position to provide insights for this process. 3) Use simulation software and develop the algorithms to estimate the model. The range of assumptions developed qualitatively are used to run the simulation several thousand times. The output of the simulation presents a range of outcomes that occur most of the time. This range of outcomes represents the envelope of estimates over which a demand estimate is likely to occur given the current set of assumptions. 4) Test the model using historical data to validate the assumptions used in its development. 5) Different assumptions can be easily tested in the model by inputting new scenarios and re-running the simulation.

Question 63

The main benefit of simulation in forecasting demand (and supply) is :

Answer 63

> The main benefit of simulation in forecasting demand (and supply) is not in arriving at a more accurate forecast, but in developing a better understanding of what factors influence demand and supply, and what processes may be causing bottlenecks in the flow of human capital. In this sense, simulation is considered to be more descriptive of processes than prescriptive

Question 64

What is the advantage to simulations?

Answer 64

> The advantage of simulation is in providing knowledge around how demand estimates might react to changes in environmental factors, in customer characteristics, in the training or skills requirements of employees, or in any of the assumptions that are used to build the simulation model.

Question 65

Simulation models require skills in :

Answer 65

> Simulation models require skills in mathematics and logistics to develop the algorithms used, but the conceptual basis of the model and the assumptions around the building of the model require as much involvement from HR and the business line.