Hiring the Right People in a Labor Market Flush with Talent: The Importance of Valid Selection Practices

It’s about supply and demand, human-resource style. Remember when the unemployment rate in the United States was less than 5%? According to the U.S. Department of Labor’s Bureau of Labor Statistics, that describes about five of the past 10 years. That also describes times when employers spent much of their energy on recruiting talent, or wooing top performers to apply for jobs.But times have changed. The unemployment rate now sits at 8.9%, which means employers everywhere are coping with huge numbers of job applicants for a small number of job openings. Recruiting is still important, and valid selection is always important. In a labor market flush with talent, however, figuring out whom to hire from the crowd is all the more crucial—and tricky.

So what really has changed? It’s something called the “selection ratio.” Simply put, the selection ratio is the number of people you can hire divided by the number of job applicants (read more). In a tough economy, lots of people are vying for a small number of jobs. Companies that used to have 40 applicants for 20 jobs (a selection ratio of 0.5) now may have 400 applicants for 10 jobs (a selection ratio of 0.025).

In other words, the labor supply is increasing, but the labor demand is shrinking. Therefore, many employers have the opportunity, theoretically, to make huge gains in terms of human capital. This assumption depends upon a crucial, often-overlooked and taken-for-granted part of the hiring process: that the procedures organizations are using to make hiring decisions must actually work. Namely, the methods used to choose new employees must adequately distinguish people who have what it takes to succeed from those who don’t.

Industrial psychologists have investigated personnel selection for decades. Although describing the nuances of designing a valid selection system is beyond the scope of this article, a number of general guidelines hold true across most industries and job types.

1. Conduct a proper job analysis. This involves figuring out what exactly constitutes performance regarding the job position in question, helping the hiring manager to know what types of knowledge, skills, abilities, and other characteristics (KSAOs) a successful job applicant should embody.
2. Understand the legal aspects of hiring. For example, employers should pay attention only to those aspects of applicants that are relevant to the job, or “bona fide occupational qualifications.” Consult an attorney well-versed in employment law to clarify any other concerns, including compliance with the Uniform Guidelines on Employee Selection Procedures (read more).
3. Conduct structured interviews. Many hiring managers assume that they can gauge the quality of job applicants through an informal conversation and a handshake. Quite simply, this is a big myth. And thousands of pages of rigorous research provide evidence to this effect. Structured interviews should function like any test: They should have specific questions designed to assess specific KSAOs, and hiring managers should conduct them in the same way for all job applicants who pass initial résumé screening. Here’s a decent guide for conducting structured interviews from the U.S. Office of Personnel Management.
4. Consider the use of psychological tests and other selection hurdles. The decision to use these types of selection procedures depend on a variety of factors, and most organizations should consult an expert for further guidance.

Overall, a labor market with large numbers of qualified job seekers can be a very good situation for employers. It gives organizations the opportunity to build their human capital with less of a need for recruiting. But to truly capitalize on the current nature of human-resource supply and demand, employers must understand and implement valid selection procedures. Otherwise, it’s just a shot in the dark—resulting in random selection that is unfair, potentially illegal, and bad for both organizations and employees.

SPSS Syntax 101: Basic Guidelines, Variable and Value Labels

Perhaps you’re now convinced that using SPSS syntax might save you some time in the long run. Maybe you even now know how to create a new syntax file. So what do you do with that file and how do you make it manage or analyze your data?

As a reminder, this forum focuses on common analyses performed by researchers and practitioners within organizational behavior, industrial/organizational psychology, and human resource management. An in-depth explanation of everything SPSS syntax can possibly do is far beyond this forum’s scope. The focus here is on practical issues and no-nonsense knowhow to bolster your productivity.

To begin, let’s cover a few guidelines regarding the use of SPSS syntax.

• Notes: You can make notes to yourself within the syntax file about what you are analyzing, why you are analyzing it, what results you found, and anything else that may help put your analysis in context for future reference. To make a note to yourself, simply precede your note with an asterisk (*) and end your sentence with a period. That will ensure that SPSS will skip over that sentence when running the analysis you specify. It is sometimes useful to use many asterisks in a row to demarcate different sections of your syntax.
• Sensitivity: Although syntax is not case-sensitive (to the best of my knowledge), syntax code is very sensitive to every other character you type (including periods and other punctuation).

Now, let’s practice. Open a data set and create a new syntax file. Click anywhere in the syntax file’s blank box. Insert a row of asterisks, a description of the file, another row of asterisks, and a period. Save your file somewhere appropriate with an appropriately distinctive file name. Your syntax should now look something like this (click on the image for a larger view).

View fullsize

The next step is to insert code into your syntax file that will specify the types of analyses that the program will conduct on your data. Three general ways that I typically generate code are through (a) directly typing it in from memory, (b) copying, pasting, and editing syntax from another source, and (c) using the program’s “paste” function. I recommend starting your foray into SPSS syntax using methods (b) and (c), which may or may not naturally progress into method (a).

One common task when setting up your data for analysis is properly inputting your variable and value labels. This is helpful because when you run analyses, the SPSS output will include the labels assigned to the variables that were analyzed, assisting in your interpretation of the results. Let’s imagine that one of your variables is “overall satisfaction with life,” being measured by one item on a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). Forget, of course, that using a single item to measure a construct is a bad idea from a measurement perspective.

Let’s go ahead and designate a variable in your data set as having the name “lifesat.” Select the “variable view” on the bottom left corner of your data file and type “lifesat” into the upper left cell under “Name.” You can quickly assign labels to both the variable itself and to its values using the syntax below:

variable labels lifesat 'overall satisfaction with life' .

value labels lifesat 1 'strongly disagree' 2 'disagree' 3 'neither agree nor disagree' 4 'agree' 5 'strongly agree'.

To run this syntax, first copy and paste it into your syntax file. Then, simply highlight it and type CTRL+R on your keyboard. You can also select “run” from the menu at the top of the syntax file or click on the right-facing triangle in the menu bar.

An output file should appear which will simply repeat the syntax listed above. Note that you will receive an error message if there is no actual variable named “lifesat.” If you look at the “label” and “values” columns in the data set, the labels and values specified above should now appear.

This procedure is especially useful when setting up an SPSS data file from raw data, which often is in the form of an Excel spreadsheet or comma-separated values (.csv) file. For example, this is specifically how the popular online survey tool SurveyMonkey compiles data for users to download. It is especially helpful when conducting initial SPSS data file setup to list all of the variable names in one column, followed by all of the corresponding variable labels in the adjacent column, and then use the concatenate function in Excel to help create all of the syntax for you. Download this Excel 2007 spreadsheet for a quick template; pay attention to the comments in each cell regarding instructions. Note that the workbook has two spreadsheets, one for variable labels and one for value labels.

I have frequently found these basic SPSS syntax procedures to be useful in the initial-setup phase of data analysis. Future posts will discuss and explore using syntax to actually analyze your data, but these initial procedures are useful because they will aid in the interpretation of those analyses.

SPSS Basics: Getting Started with Syntax

Getting started is half the battle, especially when trying to learn a new software program or trying to accomplish a new task in a familiar one. In a previous post, I discussed some of the advantages of using syntax in SPSS instead of the software program's drop-down menus. As an astute reader commented, syntax can be very frustrating because it requires the user to follow its specifications very closely. That means that any misplaced characters or punctuation can prevent your program from running correctly, or from running at all. So is SPSS syntax worth learning for garden-variety social scientists studying organizational behavior, industrial and organizational psychology, and other related fields? Absolutely. But that's because I'm going to try to demonstrate how you can use SPSS syntax in a smart, efficient way to do the basic tasks that you may encounter on a relatively frequent basis in graduate school classes, research projects, and in providing consulting recommendations.

I'll do that through posting actual syntax code that you can copy, paste, edit, and run on your own data. That should help with tasks including computing and recoding variables, calculating descriptive statistics, and conducting analyses such as reliability analysis, factor analysis, correlations, and hierarchical regression--including mediation and moderation analyses. I'll also post some of the tips that I've learned the hard way regarding keeping your syntax and data files user-friendly.

To get started, however, I'm going to assume that you've never used syntax before. You may even be a little scared of it. You're studying or previously studied psychology, organizational behavior, or human resource management because you want to help people and organizations work and play well with one another.

No one told you'd have to be a statistician or computer scientist. But neither am I. All I'm saying is that good consulting and good science builds from an understanding of research methods; therefore, it makes good sense to figure out how to use the tools we have available in the smartest, most efficient way possible. Statisticians, computer scientists, and actuaries: You may want to search elsewhere for solutions to your specific challenges.

SPSS File Types

SPSS has three common file types associated with it. First, there's your data file. This file ends with the file extension .sav. Second, there's your output file. This file ends with the file extension .spo, and SPSS automatically creates it when your run an analysis. Third, there's your syntax file. This file ends with the file extension .sps.

Creating a New Syntax File

To create a syntax file, start by opening the SPSS program or a data set (.sav file). Then, go to the drop-down menu at the top of the screen, go to "file," then "new," and select "syntax file." A new screen should open that looks like the image below. Note that an empty data set is in the background, and the new syntax file is in the foreground.

View fullsize

The syntax file in the foreground serves as your "command center," so to speak, for your data analysis. I recommend maintaining a single syntax file for each project. For example, I used the same single syntax file for all of the analysis relevant to my master's thesis and have separate ones for other projects. Basically, the syntax file is a text document that you can save, "save as," copy and paste items into and from, and so on.

Helpful hint: If you're ever on a computer that doesn't have SPSS yet need to open your syntax file for some reason, you can open it as a text file in Notepad or Wordpad. You won't be able to run the program, but you'll be able to see it, read the notes you made in it, and copy its contents.

So now you know (a) what a syntax file looks like and (b) how to create a new one. Future posts will address next steps and general tips for using the file to analyze your data. Congratulations--you got started, and that's half the battle.

SPSS Basics: Why Syntax Beats Point and Click

Consider yourself very, very lucky: It wasn't that long ago that researchers were calculating all of their statistics by hand, without the use of powerful statistical programs like SPSS, SAS, and Excel. But just because software running on powerful computers allows anyone to spit out statistics doesn't mean that you don't have to know what you're doing and to conduct your analysis in a meaningful way. Because, after all, the numbers that your software program of choice quickly hurls back at you after a few short commands is only as good as (a) the data itself and (b) what you told it to do in terms of analysis. It seems that different academic disciplines gravitate toward different software programs, and many researchers in industrial/organizational psychology, organizational behavior, and human resource management tend to use SPSS. And that's just fine for the majority of the types of analyses we like to conduct (with structural equation modeling being a notable exception). One of the big differences between SPSS and other statistical programs is that it has a variety of "point and click" options, meaning that one can tell the program to run a variety of types of analyses without having any knowledge of the program's programming language simply by using drop-down menus. While that may sound nice to beginners, it has its pitfalls. Most notably, it encourages a haphazard approach toward data analysis that in the end leaves users wondering, "What did I just do?" So I strongly encourage anyone getting started with SPSS to quickly learn to start doing all of their analyses with syntax.

What's syntax? Syntax is what we call the programming language used to conduct data analyses in SPSS without using the “point and click” method. It's useful for a multitude of reasons, including:

• It helps us conduct our analyses in a meaningful, systematic manner
• It is a lasting record of what analyses we conducted
• If something does not work, we can easily find out why
• It can perform operations that are laborious or impossible using the drop-down menus
• It is easily transferred among collaborators

Future posts on this forum will further expand upon the basics of using syntax for analyzing data in SPSS. But for now, consider abandoning the "point and click" method. You'll be glad you did.