[1] "We are all humans in job interviews"

Which selection procedure or interview medium is the right one for a particular job profile - that is ultimately the crucial question in the field of research to which Prof. Dr. Basch has dedicated his career. "In short, my work is about using the right procedures to find out the suitability of applicants for a particular job - without scaring them away in the process." Enthusiasm for this particular area of psychology has accompanied the scientist from his bachelor's degree, in which he studied intelligence tests in personnel selection, to his doctorate, in which he researched digital job interviews, and today in his work as a professor of business psychology. "We are all humans in job interviews, both on the part of the applicant and the interviewer. At the end of the day, every job interview is a social situation - and I'm committed to psychology because I'm so interested in those relationships and how people interact." 
Job interviews, he states, are an excellent tool in personnel selection. On the one hand, this is due to the fact that they are probably the best researched of all selection processes, if only because almost every company uses them. On the other hand, interviews are not only popular with companies, but are also well received by applicants. "In this way, personal contact is established, both companies and applicants can present themselves well, and ideally there is a high degree of job relevance," summarizes Prof. Dr. Basch.

[2] Rely on your gut feeling? Better not (only)!

So are traditional job interviews the ultimate in making the right choice? No, says the expert. "As I said, only structured job interviews^INFOBOX that are conducted on the basis of a requirements analysis achieve good results - and in practice, this is precisely what is lacking." Especially in times of a shortage of skilled workers, a company has to be attractive enough for the best candidates, explains Prof. Dr. Basch. "But unfortunately, there are still far too many companies that conduct unstructured, poor interviews."

As a result, if only a handful of suitable candidates apply anyway, and if the selection process then also shows shortcomings, this can have devastating effects not only on the individual job, but also on the fundamental attractiveness of the employer. "One of the reasons more time and money is not invested here is human ego," says Basch. "Most people are convinced they can assess others well and rely solely on their gut feeling. But that is by no means the case, because we are all subject to certain prejudices and take distorting factors - attractiveness, sympathy, first impressions - into the equation."

The remedy can be precisely a structured job interview, because the more structured and standardized the process is, the better these distorting factors can be cushioned. "There are studies that show: The less autonomy you give people in personnel selection decisions, the more valid their selection becomes." In practice, of course, things don't work out quite so black and white. Basically, a distinction can be made here between so-called holistic decision-making - a more intuitive decision-making process based on gut feeling - and statistical decision-making, which is based on clear measured values and rules. The personnel diagnostician himself is not an advocate of overly mechanical procedures, but rather advocates a balanced mix of both forms of decision-making: "The famous beer question, for example - the team sits down together after the end of the interview and considers whether they would go out for a beer with the candidate, i.e. a rather holistic judgment - can be a valid addition to the statistical decision-making process if it is consistently embedded in the rule. But it must never be the only or decisive criterion. It is therefore unfavorable if one already has a statistical judgment and then modifies it holistically à la 'According to our decision rules, this is by far the most suitable candidate, but I don't like his nose, so we won't take him'".

[3] Recruiting by video call: the use of technology-mediated job interviews

From telephone interviews to interviews via video conference to (a-)synchronous video interviews: Prof. Dr. Basch has a special research interest in technology-mediated job interviews, i.e. all those formats that do not require a face-to-face situation^INFOBOX. He is particularly interested in differences in performance and acceptance between different ways of conducting interviews and is investigating the validity of technology-mediated job interviews. It is obvious that their use has increased considerably since the corona pandemic as a substitute or supplement to traditional face-to-face interviews. But how do job applicants and interviewers assess these formats, some of which have become alternatives without an alternative during the contact restrictions of the past two years, and how valid are their predictions?

[4] Prefer face-to-face? What applicants don't know, they initially reject

Before the pandemic, the research situation painted a relatively clear picture, explains Prof. Dr. Basch: applicants rejected technology-mediated job interviews and preferred face-to-face interviews. The reasons behind this are manifold: At first glance, the online interview suffers from poorer presentation options, technical barriers, and the lack of opportunity to make contact. However, another point for the lack of acceptance seems to be one thing above all: a lack of experience with this format. In one of his studies, Prof. Dr. Basch was able to show that the gap in the fairness perception that applicants had with regard to video conference interviews and face-to-face interviews was significantly greater before participating in a video interview than afterwards: "The participants did not find it as bad afterwards and not as unfair as they thought - so it is also often skepticism that prevails here. And Prof. Dr. Basch took away another interesting finding from his studies: The acceptance of videoconference interviews increased the more the pandemic with all its limitations was brought to the fore in the experimental set-up of a fictitious job interview. In other words, the subjects showed a more positive attitude toward technology-mediated job interviews when the accompanying circumstances required it.

[5] The basis of every job interview: social presence

The corona pandemic has undoubtedly played a major role in improving the acceptance of virtual job interviews, not least because a large number of applicants were able to gain experience with video conference interviews within a comparatively short period of time. In the meantime, they are therefore also much better accepted - although, as Prof. Dr. Basch emphasizes, they still don't quite come close to their face-to-face counterparts. "I have found in my studies that one construct that lies beneath interviews is social presence. This refers to the perception of the other person in the same room, but is difficult to define." That a different form of social presence comes about in a video conference - a kind of telepresence - than in a face-to-face conversation stands to reason: "If I only see a small zoom window of the person I'm talking to, but not what she's wearing, how she moves around the room, or how firm her handshake is, then, in esoteric terms, the aura that is created when I can really perceive someone in their entirety is missing - but the more I feel this aura, the more comfortable I also feel with doing Impression Management^INFOBOX; that is, putting myself in the most favorable light possible."

[6] Mixing of interview media results in unfair selection process

In his research, Prof. Dr. Basch also found that candidates in videoconference interviews generally perform worse or are assessed more poorly than their direct competitors in face-to-face interviews. On the one hand, the applicants are less able to present themselves virtually, and on the other hand, the evaluators are subject to other distorting factors than in a face-to-face interview with the competition. This means that the selection process for one and the same job must be designed in the same way for all applicants - in the same medium. "Even if only one out of 20 applicants is not on site and is to be assessed via videoconference interview, the interviews of the remaining 19 applicants must also be conducted via videoconference," the psychologist points out. "Mixing interview media within a single process leads to an unfair, disadvantageous selection process."

[8] "An AI has just as much bias as we humans do".

Wouldn't a wholly automated, quasi-humanless selection process then actually be the fairest method? And, one step further: a selection process controlled purely by artificial intelligence (AI) or a recruiting robot that asks me where I see myself in five years - are these realistic scenarios then? Hardly, says the business psychologist. "First of all, we have to state that an AI in personnel selection can be many things: for example, avatars that react to me in asynchronous video interviews, that automatically recognize my spoken words, or that additionally analyze my facial expressions or gestures in job interviews and make a hiring recommendation based on that. However, an AI can also be a simple algorithm in the sense of statistical decision-making (for example: Master's degree grade < 1.5 and among the best 10% in the selection test --> hiring). These algorithms can in turn be programmed by oneself or one can give them a data basis on which they can learn themselves (neural network). And here is the crux: the algorithm is only as good as the data I use to "train" it.

The problem here is the human being: "As long as we are not free of bias - which we will never be, because we are human beings with prejudices and subjective experiences - an AI will also be subject to bias. For example, if you give the AI data from a company that has always hired old white men, and you train the AI with just those staffing decisions, it will likely screen out young, female PoC as unsuitable - because they don't fit the mold it's been taught. Very few companies work in this area with continuous learning algorithms that continue to learn with new data. As a result, old white males continue to be perceived as the most suitable, even though the AI - based on very low probability data - has in the meantime identified a young, female PoC as suitable, who then goes on to perform in the company.

The situation is aggravated by the fact that science is lagging behind practice here to a certain extent: Although the use of AI in personnel selection is already practiced in many places, it is currently only insufficiently researched. Many organizations and companies are trying their hand in this field, for example by means of automated facial expression/gesture analysis, but whether these procedures are actually valid and what kind of Adverse Impact^INFOBOX they entail - that has hardly been investigated so far. "AI unquestionably offers great potential here, because at least it already has consistent decision rules for the setting," says Prof. Dr. Basch. "But in my view, there are still too many unanswered questions for me to recommend widespread use of AI in personnel selection."

[9] Gamble until you get a job offer? Gamification in personnel selection

In addition to the use of artificial intelligence, much of the current recruitment discourse also revolves around the topic of gamification, i.e., integrating playful elements into selection processes. Gamification has reached its peak in the form of game-based assessments^INFOBOX. The idea behind this is to increase the attractiveness of unpopular selection processes and improve applicant reactions. "We know from game psychology what motivates and keeps us on the ball," says Prof. Dr. Basch. "This can work in teaching, where game-based learning platforms are successfully used, just as it can in personnel selection." However, he points out, such gamification often comes at the expense of the actual job reference, and it is also difficult to compare individual game-based assessments with one another: "We are simply dealing with too many variables here," explains the scientist. "In one of these game-based assessments, I may receive achievement icons, but not in the other; setting A has a role-playing character, setting B does not; there is a storyline here, but not there.

Probably the biggest construction site, however, is a completely different one: What sounds like a veritable application paradise to passionate gaming pros may be just one thing to others - unfair. Studies show that certain subgroups could be disadvantaged within such a gamification setting. This is particularly problematic for applicants who have a demonstrably lower average level of computer-related self-efficacy - for example, women or older people who are statistically less confident in their computer skills. "The potential discrimination of subgroups by certain selection procedures is not only in the field of gamification, but fundamentally a question that is as complex as it is highly exciting," adds Prof. Dr. Basch.