On this episode, we discuss the relationship between leadership beliefs and leadership practices.
Tune in to hear us discuss the paper Site Managers and Safety Leadership in the Offshore Gas and Oil Industry and its survey’s findings.
Topics:
Quotes:
“If we think about the effort it would take now to try to actually get thirty-six organizations to, at the same time, want to do the same research project, may be near-on impossible.”
“I don’t think there is any particular reason to believe that people’s attribution of accidents changes with experience and leadership style.”
“Once we try to fix problems with safety by putting in systems and procedures...it’s not a case of being able to just easily build back in good leadership…”
Resources:
O'Dea, A., & Flin, R. (2001). Site managers and safety leadership in the offshore oil and gas industry. Safety Science, 37(1), 39-57.
David: You're listening to Safety of Work Podcast Episode 7. Today, we're asking the question, what is the relationship between leadership beliefs and leadership practices? Let's get started.
Hey everybody. My name’s David Provan. I’m here with Drew Rae and we’re from the Safety Science Innovation Lab at Griffith University. Welcome to the Safety of Work podcast. If this is your first time listening, then thanks for coming. The podcast is produced every week and the show notes can be found at safetyofwork.com. In each episode, we ask an important question in relation to the safety of work or the work of safety, and we examine the evidence surrounding it. So, Drew, what’s today’s question?
Drew: The question for today's episode is, what's the relationship between leadership beliefs and leadership practices? Where this question comes from is that leadership is something that every single theory in safety agrees as important, whether you're talking about how reliable are the organizations, or safety culture, or behavioral safety, or the [...], they all say that leadership matters.
What that really means and how do we examine that from a research point of view. Often, the only way to find out about leadership is either to watch it happening from the outside or to get people to talk about what they believe as leaders.
In this episode, we want to check the relationship between those two things. How much can you know about how someone behaves as a leader from what they believe about leadership and what they believe about safety?
David: It's a lot of talk in safety. We have a lot to talk about the relationship between beliefs and behaviors as important across a lot of people in organizations in relation to safety. So, what are your thoughts on this connection between what people think or want to do and what people think they actually do in the workplace?
Drew: I've got really mixed thoughts about it. On the one hand, all of our theories that drive the way we try to manage safety assume that they have this strong link that if we change the way people think about safety, that will change the way people do safety.
But then, we constantly running into this examples that don't seem to match that, where people don't have as much control over their own work, they don't have the discretion to apply their own beliefs. It doesn't surprise me at all that we might find leaders performing task that they really don't believe in.
David: Last year at safety conference, I was talking about safety logics and safety culture. I opened with the following story. I listened to this play along at home. It started by saying, "Who think safety culture is about the collective beliefs of people in the organization?” and most people put their hand up. I said, "Who thinks that it's those beliefs that drive behaviors?" and most people had their hands up.
Then, I changed [...] and I asked, "Who in this room of safety practitioners believes that lost time injury rates are a good indicator of safety performance?" and every hand went down. Then I asked, "Who here gets involved in their development, production, and the discussion about lost time injury rates within their organization?" and every hand went back up. This relationship between what people think, know, believe, and what they do in the workplace is not as clear as we might think.
Drew: Yeah, I love that story. I think there's two explanations behind it. One of them is that, even leaders in organizations don't really have much choice about the way they do their job as we'd like to think they're doing perhaps even as much as they think they do. When it comes down to it, a lot of what we do at work is just constrained by the big things like organizational policies and by little things like the day-to-day of the job.
The other part of it is all of this have got aspirations to do better. I know myself, I fall short of the type of supervisor that I would like to be. If you ask me what is a good supervisor, this is what do I spend my time on those two things don't line up. I'm not happy about it. but if I'm honest, that's just the way it is. My vision of how things could be doesn't match my visions of what I'm capable of.
David: The paper that we've chosen to talk about today is titled, Site Managers and Safety Leadership in the Offshore Oil and Gas Industry. This study was done by Angela O'Dea and Rhona Flin from the Department of Psychology at the University of Aberdeen. The research paper was published in the Journal of Safety Science in 2001. That means that the research was probably done across 1999 and 2000.
It looks like the study that we're talking about today was part of a whole number of publications regarding safety leadership, safety behavior, and safety climate in the oil and gas industry in the North Sea. When we think about the time and when we think about the scale of the survey that we'll talk about soon, people recall that Piper Alpha, the disaster in the North Sea happened in 1988. That obviously sparked a whole opportunity for a whole lot of research to happen across the whole industry and you don't know what has get those opportunities.
Drew, this is a survey-based study. This is the first time that we've talked about a survey-based study on the podcast. Most of our listeners would be very familiar with surveys in the workplace but before we step into exactly what was done in this study, do you want to give us your thoughts on survey research more broadly?
Drew: Sure. [...] understand of why researchers use surveys is that surveys are cheap. Doing ethnographic research takes a lot of time, it takes a lot of access, it takes a lot of permissions, it takes a lot of ethics approval. With a survey, we can access a lot of people at once. We can put it online, we don't even need to hunt them down individually, we just advertise the link and we get the responses.
The advantage of that from a genuine quality point of view is that there is quality in numbers. When we access a wide range of people, we can be more confident that we're not just hitting individual opinions, that we're finding trends and patterns that are true for a broader community of people. That [...] of community is often ones that we don't have access through the interviews. They're people who wouldn't have time to sit down, who can't give us access to the organization but can give us access to their mind through our survey.
The big challenge with surveys (and this is really the reason why we're using a 2001 study) is that matching the research question to the survey is particularly difficult. We have to have a question that can in fact be answered through a survey.
There are lots of things that we [...] and we can’t understand people's behavior through a survey. We can’t understand what people are truly thinking through a survey. All we can do is understand what people are willing to tell a researcher in response to a particular question. That makes the design of the questions and the way which we analyze them really important in ways that a lot of researchers don’t fully appreciate. They tend to put out surveys that are seeking to collect factual information. Surveys are just surprisingly bad at collecting facts. But if they design them well, they're really good at understanding beliefs.
David: Yeah, I recently reviewed a paper about the influence and they're asking a group of people the effectiveness of their influencing tactics. That's one of those questions that you actually need data from the people being influenced not just people trying to influence to know whether it's effective or not.
That's why those examples where, certain questions, you can answer with the survey like beliefs and certain things about how something works in the world. You're not going to get the data that you want to have out of the survey.
Drew: [...] talk about this particular paper. Do you want to take us through the details of who were surveyed and why they're asked particular questions?
David: Yes. I like this paper because it was such a large sample across the whole industry in the geographic region. We don't see too much of this scale of research these days. They have 200 individual participants who were offshore installation managers. That's the name in the oil and gas industry for a Site Manager who manages an offshore production facility or drilling rig.
These 200 offshore installation managers worked across 157 different installations and 36 different organizations. The operations were on the UK Continental Shelf on the North Sea, the Scottlands are in the North Sea. That was a good, large study.
If we think about the effort it would try to take now to get 36 organizations to, at the same time wants to do the same research project, you're maybe near impossible. Like we said, there's a lot of motivation in the industry and the region at the time to actually figure out what was going on in safety.
Drew: And the thing to look at when you're looking at a survey paper is not just the raw number of people that reported back but what sample those people are selected from. 300 responses of construction workers in Australia would be meaningless. It's a large number but compared to the total number of construction workers, it's miniscule.
You need to know whether those are selected from one or two particular companies, one or two particular sites, or one or two particular types of construction. In this case, we have a good idea of how many people there are at this level within the oil industry. When we get a large sample of reasonably senior people, we've got quite a confidence that we have, in fact, covered the full range of people and their opinions. It's unlikely to just being closely selected from one or two companies or one or two installations.
David: Yeah. It's exactly true. I don't recall the exact numbers in the paper, but I think there was about 350 or 360 people in the total population. This sample is 60% of the total sample of people in that population.
Drew, do you want to tell us a little bit about the hypothesis? Being a psychology study, one of the good things about psychology study is the discipline around setting research aims, setting research hypothesis and then, setting up to test them a bit, a bit like we might see in applied science but maybe we see a little bit less of in safety. Do you want to talk about what they were trying to do?
Drew: Yeah, sure. You say it's a good habit. I, personally, find it a bit of an irritating habit to reduce every research question to a hypothesis. As a general guide, when you're looking at something which is qualitative, you have a question like your how or why.
When you're trying to produce a quantitative model, you have a hypothesis that says, "Is this particular claim or suggestion true or false." In this case, they've turned all of their questions into hypothesis. Some of them matches, some of them doesn't really. Let’s look at the overall [...].
They're trying to investigate three variables. The first one is the level of experience that someone has. The second is the style of leadership. This is self-reported so it's really what they think they are as a leader. The third one is the attitudes to safety and some of the revealed behavior. We'll get into exactly how their major behavior in a survey a bit later on.
I want to say the relationship of these things as you get more senior, does that change how you live? Does that change how you behave? The second thing is just to understand what managers thought good leadership and good safety look like. They didn't turn this into a set of hypothesis that they didn’t go to tested. The idea is that their evidence will either say yes or no to each of these hypotheses.
David: Let's talk about these hypotheses and the questions that they were trying to answer is yes or no. We're largely informed by their literature review of previous studies and pretty much previous studies that were done in relevant industries within the 10 years prior to the study.
The first hypothesis that they believe was that more experienced managers will have a more participative leadership style. Participative style will be one where the leader is more willing to engage the work, he's in decision making, and the example of more having an open door policy for communication and involvement of the workers.
The second hypothesis was that more experienced managers and those with that type of participative leadership style are more likely to attribute the cause of accidents to job-related factors rather than putting the course of accidents on the person or the worker who had the incident. That was informed by more experienced managers who think more broadly about accidents and the more participative style of leader would have more open and understanding relationships with the workers and would seek to find other reasons for why things went wrong.
The third hypothesis was that more experienced managers and those with the participative style will be more positive about their own ability to develop and maintain a positive safety climate. Those that have been managers for longer and knows that had a more open style that we would normally associate with the more mature or more positive safety climate would rate their ability to create that environment higher than those who are less experienced and those less participative in their leadership style.
The fourth hypothesis was that managers will identify behaviors consistent with a participative leadership style as best practice in safety leadership. When I ask about, "What does good look like?" they would talk about those things that are consistent with the participative and open style of leadership.
The fifth hypothesis was that managers perceptions of the outstanding safety issues that are still trying to be addressed by the industry are likely to relate to human behavior and worker motivation rather than technical and procedural issues. They form this hypothesis based on all the work that happened over the previous 10 years from regulators and companies in the oil and gas industry in the North Sea that have taken a lot of effort to resolve technical, procedural, and compliance requirements. They've formed this hypothesis that the site managers were going to be conditioned to respond that the remaining outstanding issues to be resolved were now the worker motivation and worker behaviors.
Each of these five hypotheses, based on existing research or what their witnessing in the industry at the time, and then they went out to go about testing them.
Drew: Am I jumping here and saying a little about how we then design a questionnaire? There are two broad approaches we can have. One of them is very literal, which is we directly ask the research participants the questions that we want answers to. We're relying on the fact that they will understand the question in exactly the same way that we are trying to answer it, that they will be honest and direct, and we can take their answers at face value.
Unfortunately, we know from lots and lots of experiment surveys that those two things are almost never true. It's really hard to check, has someone understood this survey the way you have? I'm sure we've all had that experience when answering one of these questions.
You do agree with the following statement on a scale of 1–5 and we've really want the options six, seven, and eight. Option six is I don't understand what you're getting at. Option seven is I do understand but I think you got to load an agenda here and I actually want to send you the opposite signal. Option eight is, I'm somewhere in the middle but not because I'm in the middle but because I disagree with the premise of the question. That approach can be dangerous.
The other way we can do it is we ask people questions that drew their answers. They don't tell us the answers to our questions but they reveal the answers to our questions. That's where a lot of the most subtle psychology instruments that I designed to do. That's always why they get a psychologist involved if you want to ask a psychology survey because they really understand how to do this stuff well. Don't [...] from first principles.
David: Another thing that happens in this type of research is people just go out and try to find instruments or set of questions that are near enough to what they're trying to understand. What I like about this paper is actually put a lot of effort into designing a questionnaire that was tailored and tested to be specific to the questions and the participants in this particular study. Do you want to talk about how they did that questionnaire design process?
Drew: Sure. I'll just take you through the basics of steps. The first thing they wanted to do is understand what their participants experience or [...] with questionnaires. That may seem a bit surprising given everyone nowadays is used to having through this questionnaires. It still is an important step to understand what other surveys are going on in an organization. You think we’re just being asked to do three separate safety climate surveys over the past three years. Giving in another survey that looks like a safety climate survey, they're going to interpret it based on the experience of the questionnaires they've already done. You need to understand what questions that are used to answering and how they tend to answer them.
Next thing is design the questionnaire based on literature. This is using the literature to find the existing questions and what those questions mean. Also, using the literature to link concepts and questions together. They then give this a review by choosing your safety executives. That's partly just a research conducted to make sure that the organizations have approved it, but it also checks to make sure that the language that's being used is appropriate for the industry. Different terms mean different things, different people. If the industry is used to using one term to something, giving them a totally different term is going to be confusing.
If I put out a survey that says, "Do you do Take 5?" wherein this particular industry is not called a Take 5, it's called a Prestart Risk Assessment or a mini moment, then they’ll say you would say no to the fact they do Take 5 is going to be meaningless. You need to know what language they use.
They didn't send out a test survey. The sent the questionnaire out to 10 site managers and they didn't just get the responses back. They sat them down and gave them interviews to talk about the questionnaire, to check how they understood each question, why they were given the answers they did to make sure that people were responding in the way that the survey designers intended. That result in the [...] of the survey which was, again, approved by the senior safety executives. All these steps are quality control even before the survey goes out. It's too late once you got 500 responses to realize that one of your questions didn't work.
David: Yeah and we shouldn't underestimate the effort that goes into designing a questionnaire, but if you really want data that's aligned to the questions that you're asking, then you need to invest the time upfront to get the questionnaire right.
Drew: Yeah. You and I know that I've got my own safety [...] questionnaire that I've been working on. It's currently sitting in, I believe, version 34 and it still doesn't work despite extensive testing. That's the challenge. Just because you got an idea, just because you think you got some good questions doesn't mean that respondents are going to respond in a way you thought or in a way that is useful [...] get out the survey.
David: Yeah. So, let's talk about the exact questions that they ask. I think there's some you seen in knowing this questions and even without the findings, our listeners might be able to take some of these questions or ideas for how they might get data in their own organizations.
This is an existing questionnaire for leadership style. One of the good things in the paper is to actually list all of their questionnaires they considered and why they dismiss certain questionnaires in relation to leadership style. I went back to 1958 leadership style questionnaire because it was short, it was clear, and it was directly on to the question around direct versus participative or mainly direct versus mainly participative styles of leadership.
They did that questionnaire for the first hypothesis. Then, they founded another study that had been done which identified the top 10 causes of incidents. They just ask the participants to list the top 6 out of the top 10 causes of incidents, which enable the pretty simple analysis process to happen.
They then ask leaders on a scale of 1–7, how easy or difficult they felt it was to create certain aspects of a positive safety climate. For example, one of the items was, how easy or difficult is it to get people to openly report near misses. So, 1 is very easy, 7 is very difficult and you pick a point in between.
This is trying to understand the leaders’ belief of the ease or difficulty for them to create a positive safety climate. Then obviously, matching the hypothesis back to experience and leadership style.
I then ask three open question and these three questions I really like. The first is, what is the most effective thing you have done to improve safety climate on your site? The second, what is the best way a leader can demonstrate commitment to safety? Thirdly, what is one piece of advice you're giving new site manager to impact safety on site?
Drew, before I go on to the last bit of questionnaire, what do you think about these three questions 20 years later?
Drew: I love these questions. I think there's a real tendency when we're designing surveys. To create questions that are easy and boring for people to answer. We put this scale of 1–5 like it scales because people know how to answer them, they can do them quickly, we get a high completion rate. But now, we never have to stop and think. They don't reveal much, they just working within categories that we have provided people.
These questions are really cleverly designed to reveal the way people are thinking. We don't actually want to know what's the most effective thing people have done. What we want to know is that by their choice of what they think is most effective, they're revealing the way they think about safety. Even when they talk about what advice they give a new site manager, they're revealing something about the way they think. We don't know what exactly what they're thinking, they're maybe thinking imagining a particular new site manager, it maybe something about their own values, but at least they're generating something that is not there build into the survey design. It's something brand new that we can look at and interpret.
As long as people are willing to fill out these open questions, there's a lot of interesting stuff that can come out of them. Often people think, "Well, open questions just means it's the end of my 1–5 scales." I'll ask the question, "Is there anything else you'd like to tell me?" These are very focused open questions that reveal a lot.
David: Yeah. They continue that for the final question through the last. The final question was just asking the site managers to list the five main outstanding issues which still needs to be tackled in a site and industry. Those last four open questions solicited a total of something about 670 free text responses.
You say that the question is with categories, tick boxes, and scales are easy to complete. They're also a hell of a lot easier to analyze than sitting there going through 670 individual free text fields and trying to do automatic analysis of it. Credit to the researchers for analyzing the necessity for needing to get the open answers.
Drew: Yeah, and the easy to analyze is called for doesn't give you particularly interesting answers either. If you want to understand safety climate, you don't ask people a bunch of questions that you already know what the two options are. That can help you make a safety climate if you already fully understand it but it tells you nothing new. You're able to find something new and you ask someone a question that you don't know the answer to. That means not giving people a choice between two options but giving them freedom to respond.
David: Exactly right. I think once you get all of the questions that you think you want to click your data, this paper particularly goes, "Am I going to get people to do this? Is this going to be too onerous?" I like the fact that this total questionnaire only contain 26 questions across trying to testify hypothesis. That's relatively short. I know that I can think of at least two times in the last couple of months where I've gone, "Okay, I'll do this survey." Got to about page 3 or page 4 and realized I was only part way through when, "Ah, this is getting way too long. I'm going to dive out of this now."
Drew: Yeah. Over then, I did lots of surveys myself. On the temptation on the other side, you're thinking, "I'm doing all this work to put out a survey and then, I get 300 responses. Oh gosh, by just making these little extra question, I got a whole new research paper over these one extra question."
On [...] is one other question, one of my PhD students is doing research project as well and it would not really help him to get the answer to this question as well. Before you know it, you've doubled the size of the survey and gone a quarter of the number of people actually responding properly to the survey. It takes real discipline to say, "I’d love to ask more questions but I'm going to keep it simple and get the answers."
David: Yeah, exactly right. That was the methodology, a custom-designed questionnaire to match the hypothesis that the researchers were trying to understand and then, a total of 26 questions including both scale-based questions and open free text questions.
Let's talk about the findings. I might just laid off here and just dive in when you want to comment in relation to the findings and we'll talk about the findings in relation to the five hypotheses.
The first finding was that, experience is not the dominant factor in determining leadership style or attitude to safety. This idea or the hypothesis that was that more experienced leaders would have a more leadership style was in that true-false way, based on the data, determined to be false that there was no statistical relationship between years of experience and leadership style.
Drew: Yeah. I like that one because researchers tend to ask questions in the direction in which they’re expected. These research clearly went in and in fact, in literature, it gives the reasons why they were sure that as people get older, they would have a more participative leadership style. I like the fact that the very first one is just, this was true but no.
David: And here we go. Hypothesis two, not sure as well. So, the second hypothesis was that a more experienced leader and a leader with a more participative style will attribute safety issues less to personal relative factors.
Like I said when I explained the hypothesis, if you're more open, you've got closer relationships, more information coming out a few people, and if you've got more experience, you've seen more things, you've heard about more, or witness more complex issues, then you're not going to straight away go, "Okay, the incident was the worker's fault."
But I'm not sure if the three most common causes of accidents that were identified by site managers where number one, the worker not thinking through the job. Number two, worker carelessness. Number three, worker failing to following the rules. No difference between more and less experienced leaders, participative, or directive leadership styles.
Drew: That one surprises the researchers but it doesn't surprise me at all. I don't think there is any particular reason to believe that people attribution of accidents changes with experience and leadership style. It tends to change with what people feel they have control over or with their ability to separate themselves out from the causes of accidents.
People would love to believe the next [...] will never happent to them and one of this is ways to believe that is that to believe it happens through carelessness or failure to follow the rules or for being dumb because then they can say, "I'm not any of these things so I'm safe." That's a pattern we see over and over again and yet, I don't find it particularly surprising or interesting that it brought up again here. It wasn't what the researchers were expecting.
David: Yeah. I mean, I was a little bit surprised because I just reflect on my personal experience throughout my career and as you get more experienced, as you read more things, and see more things and you evolve the way that you think the world works and I've definitely done that myself but I think, maybe the offshoring installation managers don't have access to the learning opportunity that I've had to try. I don’t know whether my change in idea is being because of my experience or things that I've read or learned, and those are two different things.
Drew: So, you've become older and wiser, David. It's not necessarily a correlation between the two, as some of us get more reactionary and set in our ways as we got older.
David: I'm not sure. That's a bit too much of a complement. Maybe someone did said to me, "You can have 10 years of experience so you can have 1 year of experience 10 times."
Drew: Yeah.
David: Hypothesis number three. The finding here was that there was no strong relationship between the experience and leadership style and perceptions of the difficulty in creating a safety climate. This idea of the more experienced someone is in a leadership role and the more participative their leadership style, then they're going to report that they find it easier to create a positive safety climate.
That makes rational sense. The opposite was found to be true. It was actually found that less experienced leaders and those with the directive style, actually claim that creating a positive safety climate was much easier. The researchers concluded that because of the way that the sample was cut that these inexperienced or less experienced leaders and more directive style of leaders were over estimating their ability to influence or motivate the workforce.
I had to me, the reflection here about the politics of safety. What you have here is a situation where the more experienced leaders and the ones with an open style of leadership that you would more align with a positive safety climate where then saying that they were finding the task of leadership harder and I think if you start at getting that data within your organization, it would really strain the way that you think about the performance of people and their role.
Drew: This is where we need to be really careful about the limits of people self-reporting their own behavior and their own success. We could take this at face value and say, "Everyone who is very directive in their leadership is confident that it works. Everyone who is very participative is much more uncertain, so let’s listen to the people who are more confident.” Or we could do what the researchers did and interpret it backwards and say, "They all saying that it works but that just means they're over confident."
I don't think the data really gives us a “have to know” which of these cases. We don't know whether this less experienced leaders are more successful or whether they are over confident. So, they can't tell us the answer to do that. All they can do is tell us they think that they are more successful where's the more experienced, the more participative ones think that it is harder.
David: Yeah. I think one of the conclusions here was that no one says it was easy and all leaders says that they found it difficult and maybe it's just the case of the less experienced and directive leaders were unconsciously incompetent and maybe the experienced and participative leaders were more consciously incompetent and it’s reflected in their ratings. But all leaders reported that they found it very difficult to create a positive safety climate.
Number four, finding number four was although, managers are aware of best practice leadership, they do not always act in a way consistent with this. I remember those three open questions where the researchers ask, what's the one thing you've done to improve safety climate and what's the advice that you'd give to a new leader and what's really important?
The thematic analysis of those results identified four key areas. One was visibility, we talked about visible leadership all the time. Second was relationships, supervise our worker relationships and fostering worker-worker relationships. The third was workforce involvement, engaging and involvement in conversation, decision-making, ownership of safety practices. The fourth was proactive management, so getting on top of issues before they become problems.
Four practices and factors that we would now gloss over and say, "Yeah. That's all part of it." But this leaders had all consistently identified these four things as priorities for best practice safety leadership but then, they reported that this wasn't the way that they lead in their own roles.
Drew: Yeah. That's where the sneaky questions come in when we ask them what's the piece of advice that you give to a new site manager? Having said that we practiced safety leadership is visibility. Number one piece of advice is not be visible. What are the main causes of accidents? It's not failure to get out of relationships, it's people stuffing up.
This is where people's [...] behavior, what they say that they believe, and they reveal behavior, what they say are the top issues or what advice would they give come into conflict. It's a clever way of getting two separate things measured by the same survey.
David: Yeah, I think this hypothesis and the findings here just reveal that the title of this podcast really that there is a gap and there is an unclear relationship between leadership beliefs or leaders belief about safety and then, leaders practices in relation to safety.
The fifth finding area was about, what are the open issues or areas of improvement within safety, within new installation in the industry and there were four themes that was extracted that was reported openly by more than 30% of respondents. One and three made a comment in relation to one of these four things.
The number one area was that 44 of respondents said that they need to get work as more involved in safety activities. They need to get workers to have more awareness of safety and take more ownership of their own safety. Clearly, if we can think about what these leaders reported as the cause of accidents, then clearly we can see how many of them would say or at least half of them would say that we need our workers to take more ownership of their own safety.
The second was in relation to just the general workforce competency that has killed people on working in the industry. [...] third, go over the fourth, because I want to talk a little bit more about the third in a minute. One with 31% of respondents was the need for standardization of safety culture. This was actually more about behavioral safety and if we think of the time what was happening throughout the 90s, this was specifically leaders saying they needed their people to make safety a habit, not a compliance issue.
The one that I think we might talk about a little bit was this need for what the researchers think the harmonization of procedures across the industry. What was underneath that, what this was really about was site managers requesting or playing for a simplification and then getting back to basic seeing safety.
They were saying that they have been overwhelmed in the last 10 years, and since Piper Alpha, overwhelmed with the initiatives, legal requirements, and procedures, and they were throwing their hands up and saying, "It's all too much, we need to simplify safety. We don't need to keep issuing new procedures. We don't need to keep doing new initiatives."
This was probably one of the first rock-back-on-the-chair moments through since we've been doing the podcast where I thought it was only a year or two ago that we wrote the paper on safety [...] and theorize around safety [...] as a new phenomenon, and here's 200 site managers 20 years ago telling us that it's already a problem.
Drew: I think we can certainly see in this paper the roots of a lot of the modern problems in safety. It wasn't that the managers, 20 years ago, were experiencing modern day levels of demands from certain management systems and procedures. It was that things were already overwhelming then, and what we've done ever since is to make it worse. I think we really need to link that with the other conclusion of this paper, this gap between how leaders think leadership should be happening and how they find themselves able to lead.
One of the biggest constraints that they have is the safety initiatives. The systems they’re required to manage, the compliance they’re required to manage. Those things are at odds with their espoused values of leading through visibility because they're in the office doing paperwork, doing relationships because they're required to enforce some clients.
Once we try to fix problems with safety by putting in systems and procedures, that it's not a case of being able to just easily build back in good leadership. Those are the things that were keeping this leaders from keeping the way they would like to lead.
David: Yeah, I agree. So, before we talk about the practical implications, I thought it would just be worth providing the conclusions that the authors raised at the end of their paper. Do you want to talk through those conclusions that the office provided?
Drew: Sure. Here are four bolt points, basically, in the author's own words. This one is this classical more research required statement. More research is required to understand the relationship between site managers and safety climate. That was interesting. Only very recently that we've seen a big [...] leadership and safety culture and the focus for [...] senior leadership.
The second one is the idea that managers are already keenly aware that they have a bigger role to play in safety climate and that they already know that the best way to achieve it is through participative leadership styles. This isn't a gap in beliefs or understanding.The gap is translating that knowledge into practice was really, really difficult for them.
The final one was this one you already mentioned that less experienced and more directive managers tend to overestimate their own ability to promote a positive safety climate. Then, they're doing to take away the final sentence of the paper.
David: Yeah. It's always an effort when you're writing research papers to get a title write, to get the abstract right, to get the opening sentence and the last sentence of the paper right. You spend probably half your time on those four elements and half your time on the rest of the paper.
The final sentence of the paper concludes that there is still some way to go in developing the right environment of optimum safety performance. Drew, you said before we jump on the podcast, you could almost write that on the end of every single safety sized paper that's being published in the last 20 years.
Drew: Indeed. So, let's move on then to how do we develop our environment from optimum safety [...]. What are our practical takeaways?
David: Yeah, let's go with some practical takeaways. Some of these, we haven't talked about yet but that was underneath the data. The first practical takeaway I had was, is there a real difference between owners’ and contractors’ beliefs and practices for safety? It’s a bit unfair to just talk about the practical takeaways because they haven't talked about it through the questions or findings, but mostly in the oil and gas industry of the 157 installations, there was a mixed of production platforms and draw rates.
The way the industry works is that the production platforms are generally operated by the [...] or the clients and then the drilling rigs are operated by contractors. This was reported as a surprising subfinding of the paper where there was no discernable difference in any of the questions in relation to drilling safety managers working for contractors, or production installation manager working for clients.
The practical takeaway here is that, for me, reading the paper anyway was this generally accepted and communicative view that we have among client organizations that the client organization isn't the problem, the contractor is. More specifically, for those listeners who are in the oil and gas company, you might also regulate here the view that operations or productions aren't the problem, the drilling team are.
On both accounts, I think this study, at least for me anyway, said step back and ask, why is your organization or your industry think that? Don't answer because of injury rates because we know by now in episode seven that injury rates aren't the greatest way to judge safety, but step back and ask whether or not you're perpetuating something that you don't really have good factual answers for.
Drew: The second takeaway is that in this study, the non technical issues could be seen as a priority for improving safety in workers, in particular, leadership communication and employee motivation. That equity sounds like this is something that's just we grown on about the importance of non-technical issues.
Remember that this is a study of what leaders believe. The interesting finding here is that people already care deeply about these things. They want to be better leaders, they want to have better communication, they want to be motivating their employees. It's not the belief, it's not that they need this way, it's not that they need telling, "Be leaders, be communicators, be participative." It's they're finding it hard to do that given they designed up their jobs and the situations they're experiencing. Rather than telling people to do this better, working out what we need to do to change and enable them to be able to do what they already want to do is important.
David: Yeah. I agree with that. It's a bit related to number three. The number three practical takeaway was in this study, 57% of the leaders prefer the directive style of management even though all of the respondents unanimously reported that best practice safety leadership is a more participative style and there was no clear link between experience or other factors in relation to leadership style.
If we think in our organizations that we want more participative styles of leaders, then perhaps the only way to do that is to find someone who prefers to operate with the participative style of leadership. This study shows that even if people know and believe that it's the right way to lead, it may not be the way that they either do or prefer to lead. Some takeaways here for your recruitment and your appointment of leaders with the leadership style that you want. And if you are in an organization that you want to want to have a more directive style of leadership, then design this through of the reverse.
Drew: I think that applies not just for recruitment but also for internal selection and promotion. We tend to select people based sometimes on their own job effectiveness or on their effectiveness of their immediate team, and that short-term effectiveness can come from just being a very [...] type person or a very task-driven type of person and not having participative leadership. If we want our leaders in the organization to be [...], we need to notice that, we need to encourage it, we need to reward that, and we need to elevate those sorts of people.
David: Yeah, you're right. The fourth one practical takeaway was, for our listeners, is, what do you and the people in your organization belief causes accidents? What are the accident causation logics that is shared in your organization? This study, of all of these 200 managers at the time in this industry, in this part of the world, anonymously aligned and agreed that the causes of accidents were workers not thinking through the job, worker carelessness, and worker non-compliance.
If those are the logics that are shared and held in relation to accident causation, that's where the supervisory effort, that’s where the organizational effort is going to try to improve safety performance. Now, we know that this is a set of information from one group of people and I'm sure if we survey all of the workers and all of those installations that they might not have come up with those same three things, but it’s worth all of our listeners taking some steps to understand what are the shared accident causation logics that play in your organization? What's your role in reinforcing or reshaping those logics?
Drew: And then the final takeaway is, how much are we really helping, enabling, and equipping leaders with the skill set that they need to create a positive safety climate? A big part of that that comes out of this study is just how realistic are we about what we're expecting?
In this study, it wasn't that people were saying they shouldn't believe in safety climate or they didn't think that had a role to play, it's that they said that it's very difficult to develop and maintain that climate. The more experience people had, the more they had participative style, which might be [...] that we're trying to do it, the harder they said it was.
David: Yeah. With researchers in safety and practitioners in safety over the last 20–30 years where [...] what safety climate and safety culture are, yet we have fairly straightforward expectations that our frontline leaders, our middle and senior managers are just able to create this and that's a good reflection for me in this paper. How much effort do we put into actually helping, align, enable, and support leaders to be able to create that? You can't really blame a leader for not creating a positive safety climate. What you can do is help a leader with a set of capabilities and set of activities that you hope will get there and then, judge them and their ability to do that and leave like that.
Drew: The final takeaway is just really about the nature of these questionnaires. I think we can take away some [...] interesting data we've got by asking some cleverly designed open questions as opposed to administering engineering safety culture or safety climate questionnaire.
I got some good feedback about what was considered to be good practice as well as some good information about your current safety gaps across the group. [...] is used for this research. Just imagine how useful those 670 individual comments would've been as something that you would feedback the safety team.
Again, this is what people are thinking, this is what people are experiencing, this is what people need help in fixing.
David: Yeah. I like that. I think we have suggested to our listeners that ethnographic approach is really useful for collecting data but in reality, for a number of our listeners that are working across multiple operations, multiple countries means they can’t always immerse themselves everywhere and every operation to do the ethnographic style of data collection, but the alternative is having a short set of 3–5 open questions. They can get information from across their organization that can be used to inform how they direct organizational effort to support people who are at the frontline, managing at the point of risk. I really like that as a supplementary professional practice approach.
That’s it for this week. We hope you find this episode thought-provoking and ultimately useful in shaping the safety of work in your own organization. Send any comments, questions, or ideas for future episodes directly to us at feedback@safetyofwork.com.