The training illusion: why watching the work is not being ready to do it

Editorial

The badge and the work

An organisation can prove that training happened without proving that its people are ready to work independently.

The attendance list is complete. The slides have been shown. The new colleague has watched an experienced person execute a procedure and has received a document to read. A manager may now be tempted to classify the training as complete. Yet the real work does not unfold like the demonstration. Instructions are interrupted. Priorities compete. A condition appears that the slide deck never described. The colleague must decide whether to continue, pause or ask for help. This volume studies that gap. Training activity is not the same as retained learning; learning is not the same as transfer to work; a good first performance is not proof of competence across conditions. The science is more useful than a slogan: it describes the trainee, the design of practice, the support of the workplace and the measurement required to know what changed. The author has worked with last-mile ramp-up teams and with Front Office and Night Audit training materials. During an early hotel training night a system update limited practical learning. He later organised SOP guidance, images and checklists in Front Desk Control. Those verified categories do not establish that any particular training programme failed or that the tool improved measurable outcomes; every composite scene below is labelled. The editorial test is simple: what could the trainee demonstrate, in an authorised setting, without being prompted; what exception could they recognise; and what support would be available when the real shift begins?

Stiven Janaqi, Editor

How every chapter is built

1. The living scene. A shift, a ramp, a desk at 03:00: one piece of work followed as it actually happens, with the hand-off where it breaks. Scenes are constructed from real categories of work and say so.
2. The conceptual and scientific breakdown. The mechanism behind the scene, named with the model that explains it, and the research that tests it: the study, its population, its year and its limits, before it is applied anywhere else.
3. The operating protocol for action. What a working team can try on Monday: fields, questions, a cadence and a count, each with an owner, a next step and a return of information to the people who do the work.

The reading contract

• Field. Real categories of operational work, no invented personal incidents.
• Evidence. Research setting, sample, outcome and limits stated together.
• Action. Readiness is claimed for a task, from observed work, never for a person.

Research note

Training evidence needs a denominator

The studies in this volume were conducted in different settings and evaluated different outcomes. They do not provide a universal readiness percentage.

Baldwin and Ford (1988) framed transfer as both the generalisation of learned material to a job and its maintenance over time; their review is a synthesis, not an experiment proving one optimum method. Arthur and colleagues (2003) reported separate effect sizes for reaction, learning, behaviour and results, from different collections of evaluations, not one cohort. Taylor, Russ-Eft and Chan (2005) reviewed 117 studies of behaviour-modelling training. Keith and Frese (2008) analysed 24 studies with 2,183 participants of error-management training. Blume and colleagues (2010) synthesised 89 empirical studies of transfer. Aguinis and Kraiger (2009) review the benefits of training for individuals, teams, organisations and society: one more reason not to mistake a completion count for an organisational outcome.

• Reported. Publication, population, outcome, effect size and design.
• Interpreted. An operational hypothesis cautiously transferred to other settings.
• Not claimed. A measured improvement in the author's workplaces without evidence.

The numbers identify reviewed studies or participants. They do not establish how many new drivers were ready after a course, how often hotel trainees made mistakes, or whether a private training tool reduced incidents. The business examples illustrate mechanisms and generate local questions; they do not launder research findings into unsupported employer claims.

Our reading

The decisive question is not whether the training was delivered. It is whether the work can be done safely, accurately and with appropriate support.

Baldwin and Ford (1988); Arthur et al. (2003); Taylor et al. (2005); Keith and Frese (2008); Aguinis and Kraiger (2009); Blume et al. (2010).

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Winfred Arthur Jr., Winston Bennett Jr., Pamela S. Edens & Suzanne T. Bell, Journal of Applied Psychology 88(2), 2003; Paul J. Taylor, Darlene F. Russ-Eft & Daniel W. L. Chan, Journal of Applied Psychology 90(4), 2005; Nina Keith & Michael Frese, Journal of Applied Psychology 93(1), 2008; Herman Aguinis & Kurt Kraiger, Annual Review of Psychology 60, 2009; Brian D. Blume, J. Kevin Ford, Timothy T. Baldwin & Jason L. Huang, Journal of Management 36(4), 2010

Chapter I: The first shift alone

Layer 1 · The living scene

The first shift alone

A trainee can repeat a procedure beside an experienced colleague and still be unprepared to judge an unfamiliar exception.

Constructed last-mile scene, not a personal incident

The loading area is preparing for departures. A newly trained colleague knows the planned sequence: check the assigned work, confirm the expected load and move to the next step. During practice, a trainer stood nearby and answered questions before uncertainty became visible. Today, on an independent shift, the colleague sees an item whose status does not match what the work area seems to show. There are at least three possible responses. The worker could guess and proceed. The worker could stop everything without identifying the correct owner. Or the worker could recognise a defined exception, contain only the affected task and contact the authorised role. The appropriate response depends on the real procedure, which this illustrative scene deliberately does not reproduce.

This is not proof of a mistake at an employer. The author has documented last-mile loading, missing bags, new-driver ramp-up and standard-work challenges. The specific worker, clock time, discrepancy and outcome here are invented. The scene isolates one training question: did anyone ask the new colleague to demonstrate the response to an exception, not just the ideal sequence?

Our reading

The first moment of uncertainty is where a certificate and a capability may part company.

Constructed illustration anchored only in documented last-mile work categories; no internal operational procedure is disclosed. The essay named below says why “watch how I do it” is not training.

More in the essay: “Watch how I do it” is not training

Layer 2 · The conceptual and scientific breakdown

Attendance is not competence

A training record answers who attended. A readiness decision must answer what the person can do and under what conditions.

The gap starts with a confusion of evidence. Attendance shows exposure to an activity. Recall shows whether a person can explain a step. Guided performance shows execution with prompts and protection. Independent demonstration gives stronger evidence for a bounded task. Transfer asks whether the skill survives when the original trainer, setting and script are gone. Baldwin and Ford identify generalisation and maintenance as the central conditions of transfer. The implications are practical: define the task before training, demonstrate it in a relevant environment, and observe what the worker does once ordinary support replaces the trainer. A single flawless practice run is narrow evidence when the role contains different kinds of exceptions.

• Seen. The trainee watched or attended the instruction.
• Shown. The trainee performed a defined task to a standard.
• Retained. The capability was checked again in real or controlled conditions.

Readiness also depends on task risk. For a low-consequence lookup, supported practice may be enough to proceed with supervision. For safety-sensitive work, access controls or formally governed hotel procedures, only the employer can authorise the required competence checks. An editorial card cannot certify a person or bypass the organisation's standards.

Our reading

A completed training session is evidence of delivery. It is not automatically evidence of readiness.

Research frame: Baldwin and Ford, Personnel Psychology (1988). The readiness categories are an editorial operational synthesis.

Source: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988

Layer 3 · The operating protocol for action

Before the first independent shift

A release-to-work decision should be proportional to task risk, based on observation and owned by an authorised role.

• Name one critical task. Observable completion criteria and the applicable current procedure.
• Observe an assisted execution. Separate what the trainee did from what the trainer supplied.
• Introduce a safe exception. A simulation or approved training situation, never a live unsafe shortcut.
• Ask for a teach-back. What is the expected result? Which condition requires stopping or escalation?
• Record authorised supervision. The level of assistance and the person who can approve progression.
• Schedule a follow-up observation. Does performance still meet the standard after the initial demonstration?

These categories are no substitute for local qualifications. A fair assessor records not only pass or fail but what was observed, the limits of the observation and the next practice opportunity. A trainee who correctly recognises a condition that requires help may be showing competence, not weakness; one who performs the normal task quickly while ignoring an exception should not be declared ready because the stopwatch looks good.

Our reading

Do not mark a person independent because the calendar says the training period has ended.

Original editorial protocol informed by Baldwin and Ford (1988) and Salas et al. (2012); not an employer-approved certification.

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

Chapter II: “Watch me” is not a training method

Layer 1 · The living scene

When the demonstration cannot happen

The gap between watching and performing becomes visible when a learner cannot practise in the system they need to master.

Documented context and constructed reflection, no guest details

During an early hotel training night in 2026, a system update made normal hands-on learning unavailable for several hours. The colleague could explain parts of the process, but the trainee could not practise them in the usual way during that interval. This is a documented limitation of that training night; it is not evidence of a sudden unplanned outage or of unsafe hotel operations. A useful question remains once access returns. Was the missed demonstration later converted into supervised practice? Was a person asked to perform the step, describe an exception and explain where the approved guidance lived? The records available for this volume do not establish a measured before-and-after skill level, so no invented final outcome is supplied.

The author later worked with SOP documents, annotated images, checklists and Front Desk Control to make training information easier to locate. The tool is not claimed to be a hotel-approved replacement for the official system, and no measured reduction in training time is asserted. A clear reference helps preparation; reading it is not the same as making a correct decision inside the real workflow.

Our reading

If practice is missing, the training gap must be named rather than disguised by a completed attendance sheet.

Documented context: early hotel training interrupted by a system update; the interpretive questions and remedies are editorial. The essay named below tells that night.

More in the essay: The night the system went down

Layer 2 · The conceptual and scientific breakdown

Four moves after the demonstration

Behaviour modelling can help learning. The mistake is treating observation alone as the entire training design.

Showing an experienced person doing the job is useful because it gives the beginner an example of sequence, language and visible standards. The learner still has to form an explanation, try the behaviour and receive information about the difference between the attempt and the standard. When the task changes, the learner may need a principle rather than a memorised click path. Taylor, Russ-Eft and Chan analysed behaviour-modelling training across 117 studies. Their meta-analysis examined differences in learning, job behaviour and organisational results and identified design features associated with better transfer. The researchers did not argue that watching alone creates competence: the type of modelling, the practice conditions, support and performance criteria mattered.

1. Show. Make the reference performance visible.
2. Explain. State why each decision is required.
3. Rehearse. Let the learner perform safely.
4. Review. Correct, repeat and test transfer.

A manual can make the sequence repeatable, but it cannot observe how a learner reasons about an exception. A trainer can supply that judgement, yet may make the practice easier by prompting every step. The design task is to reduce prompting deliberately without removing safety supervision; only authorised staff decide which steps may be rehearsed and which remain restricted.

Our reading

The strength of modelling is the visible example. The test of training is what the learner can later do with it.

Taylor, Russ-Eft and Chan, Journal of Applied Psychology (2005), a 117-study meta-analysis; the four-step cycle is editorial.

Source: Paul J. Taylor, Darlene F. Russ-Eft & Daniel W. L. Chan, Journal of Applied Psychology 90(4), 2005

Layer 2 · The conceptual and scientific breakdown

What 117 studies can and cannot say

A meta-analysis of behaviour-modelling training differentiates learning from later performance rather than treating them as identical.

117 studies reviewed by Taylor, Russ-Eft and Chan (2005), examining behaviour-modelling training across six kinds of outcome

Effects were largest on learning measures, smaller for job behaviour and smaller still for results. This does not mean all training decays in that order; it means different outcome measures do not automatically move together. Design features such as learning points presented as rules, trainee-generated practice scenarios and support in the job environment were associated with stronger outcomes. The original studies used different tasks, populations and designs. A pooled analysis cannot say that a hotel worker needs a given number of practice nights or that a transport employee will perform to standard after a set number of demonstrations. It supports stronger design and more careful measurement, not a universal timetable.

What was measured, and what the manager must still verify

• Learning outcome. Can the learner explain or demonstrate the target skill?
• Job behaviour. Does the skill appear without training-room prompting?
• Business results. Has service quality actually improved, without unwanted costs?

Our reading

No training effect size here should be translated into the probability that one person is ready.

Taylor et al., Journal of Applied Psychology 90(4), 692–709 (2005). The verification column is editorial.

Source: Paul J. Taylor, Darlene F. Russ-Eft & Daniel W. L. Chan, Journal of Applied Psychology 90(4), 2005

Layer 3 · The operating protocol for action

Teach the reason, then ask for proof

A demonstration earns its value when the trainee can practise, receive specific feedback and explain a safe exception.

• Start with a target behaviour. One well-bounded, approved process rather than an entire shift checklist.
• Model one normal execution. Show the finished work and narrate the important decisions, not private data.
• Give a learner-led attempt. The trainee controls the sequence; the trainer observes without unnecessary prompting.
• Offer corrective feedback. The observable gap between behaviour and standard, not a judgement about personality.
• Introduce a controlled variation. Practise recognising an exception and knowing when to stop or seek help.
• Retest later in context. Another observed attempt, with the amount of assistance recorded.

For an operational guide or a Front Desk Control-style reference, each step should point to the approved procedure, its owner and revision date. No side guide may contradict the official system or store private guest or payment information. Where a task cannot be practised safely, use approved simulation and hold the authorisation boundary.

Our reading

Show, try, correct and retest. A trainee should own the performance, not merely watch the trainer own it.

Editorial training protocol after Taylor et al. (2005), Keith and Frese (2008) and Salas et al. (2012).

Sources: Paul J. Taylor, Darlene F. Russ-Eft & Daniel W. L. Chan, Journal of Applied Psychology 90(4), 2005; Nina Keith & Michael Frese, Journal of Applied Psychology 93(1), 2008; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

Chapter III: The knowledge nobody wrote down

Layer 1 · The living scene

The expert who is the manual

A process can appear documented while the decisive judgement remains in the head of one experienced colleague.

Constructed cross-shift scene, not a named incident

An experienced team member answers a recurring question without hesitation. The person knows which version of a guide applies, why one status needs verification and which uncertainty must be passed to the next shift. A new colleague has access to the written procedure but has not learned how to interpret the edge case. The expert supplies the missing knowledge in a sentence. The work continues, and the hidden dependency remains. The problem becomes visible when the expert is absent. The document still exists, but it assumes background knowledge the beginner does not possess. A team may call this a training issue. It may be a documentation problem, a design problem or both. The correct response is not to demand that the expert remember more. It is to make the part of their judgement that can safely be documented usable by others.

The author has worked with dispersed hotel SOPs, visual training notes and handovers, and has organised a personal reference tool. No specific colleague, question, shift failure or quantified dependency is documented here. The scene is an explanatory composite, not a report of a real person's conduct.

Our reading

A document that cannot be interpreted without its author is only partly a procedure.

Editorial scenario informed by documented kinds of SOP and handover work; no confidential steps or identities are reproduced. The essay named below describes how Front Desk Control was built.

More in the essay: How I built Front Desk Control

Layer 2 · The conceptual and scientific breakdown

Tacit knowledge must meet a standard

A working team holds more knowledge than its forms. The goal is to capture usable judgement without pretending everything can be written down.

A standard operating procedure defines agreed action. Workers develop additional practical knowledge by recognising familiar patterns, anticipating recurring confusion and seeing which details another person will need. Some of this knowledge is useful and legitimate. Some is merely an unapproved shortcut or an old procedure that no longer matches the current system. A manager must not preserve both indiscriminately. The training design should ask experienced workers which evidence changes a decision. The answer is checked against the approved standard, not copied directly into a public guide. Examples can show a routine case and a boundary case. Revision records identify the owner and version. New staff need a route to clarify uncertainty, rather than relying on the personality or availability of one expert.

• Extract. What distinguishes the normal case from an exception?
• Validate. Does the explanation match approved current practice?
• Transfer. Can a new colleague use it with the allowed support?

The limits matter as much as the benefits. Some knowledge is restricted for privacy, safety or compliance and stays in controlled systems. Public educational material may explain the managerial principle while omitting guest data, credentials and operational security details. An internal guide is never a replacement for authorisation or supervision.

Our reading

Write down what can be shared, verify what must be approved and preserve the boundary between them.

Editorial synthesis using the transfer factors described by Baldwin and Ford (1988), Burke and Hutchins (2007) and Grossman and Salas (2011).

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Lisa A. Burke & Holly M. Hutchins, Human Resource Development Review 6(3), 2007; Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), 2011

Layer 2 · The conceptual and scientific breakdown

Why training fails to travel

Transfer research repeatedly identifies more than the quality of the course itself.

Baldwin and Ford (1988) grouped the antecedents of transfer around trainee characteristics, training design and the work environment. Grossman and Salas (2011) later reviewed variables with more consistent relationships to transfer, including motivation, utility, behavioural modelling, realistic practice, supervisor support, opportunity to perform and follow-up. Burke and Hutchins (2007) organised the literature along similar lines while discussing limits in the evidence base. Blume and colleagues (2010) meta-analysed 89 empirical studies and reported positive associations between transfer and predictors such as motivation and a supportive work environment. That does not let a manager calculate an individual's readiness from a questionnaire score: the reviewed studies varied in how transfer was measured.

Transfer barrier, and the practical question

• No opportunity to practise. When will this competence actually be used?
• No access to support. Who can clarify an exception on a live shift?
• Outdated guidance. Who owns the current approved version?

Our reading

Even an excellent training module cannot transfer a behaviour into a workplace that never permits its use.

Baldwin and Ford (1988); Burke and Hutchins (2007); Blume et al. (2010); Grossman and Salas (2011). For a reference like Front Desk Control, the research is a reason to test whether users can find and apply a procedure, not permission to claim that an application makes people competent.

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Lisa A. Burke & Holly M. Hutchins, Human Resource Development Review 6(3), 2007; Brian D. Blume, J. Kevin Ford, Timothy T. Baldwin & Jason L. Huang, Journal of Management 36(4), 2010; Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), 2011

Layer 3 · The operating protocol for action

Turn experience into transferable guidance

A method for converting expert context into approved, maintainable learning materials.

• Locate a repeated question. Which part of a process do new colleagues repeatedly need explained twice?
• Ask the expert for observable cues. What would you look for before deciding the next approved step?
• Compare against the owner-approved SOP. Reject informal workarounds that contradict a controlled requirement.
• Write a minimal case example. Separate the normal state, the exception and the authorised next decision.
• Test with a newcomer. Can the learner explain the case without the expert translating every line?
• Assign revision responsibility. Record the source, version, approver and next review trigger.

The tool should preserve provenance: an old screenshot of a superseded interface can be more dangerous than none if it convinces the beginner that an outdated process is current. Use privacy-safe, anonymised images; never place operational secrets or live records in a public repository. Ask too what cannot be codified reliably: judgement develops from practice under supervision, and a checklist supports but cannot replace that process.

Our reading

Make knowledge accessible without turning an informal note into unapproved authority.

Original editorial guidance; research context from Grossman and Salas (2011) and Salas et al. (2012).

Sources: Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), 2011; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

Chapter IV: Competence under pressure

Layer 1 · The living scene

The standard on a busy day

A person may understand a standard in quiet training and struggle to apply it when interruptions and volume compete for attention.

Constructed peak-volume scene, not a documented case

It is a high-volume operating period. A new colleague knows the approved sequence and has completed the training activities. The team is moving quickly. Several tasks demand attention at once, and a routine exception appears. The training environment allowed time to ask the trainer what the exception meant; the real shift demands a decision before the next step can proceed. The temptation is to treat hesitation as proof that the worker is weak. Another temptation is to excuse any skipped standard because the operation is under pressure. Neither is enough. The more useful question is whether the training included task variation, whether the workload exceeded the safe operating design, and whether the employee had access to timely help. A capable person can be placed in a process that makes good performance unnecessarily difficult.

The author has documented high-volume periods, short-notice driver requirements and ramp-up challenges in last-mile operations. The particular trainee, workload sequence and result are invented for illustration. No operator performance statistic or incident is attributed to a real employer.

Our reading

Training should prepare the person for expected variation, while the operation remains responsible for conditions in which the standard can be followed.

Constructed illustration based on documented operational categories, not an account of a specific trainee failing. The essay named below describes the standard under pressure on high-volume days.

More in the essay: High-volume days: the standard under pressure

Layer 2 · The conceptual and scientific breakdown

Transfer under changing conditions

The real test of competence includes deciding which action not to take.

A procedure rehearsed only in its ideal sequence tests reproduction. A real role may require adaptive transfer: recognising when the environment has changed enough that the approved exception pathway applies. Performance under pressure therefore has at least three dimensions: executing routine steps accurately, recognising the boundaries of the routine, and using the right support when the boundary is crossed. Training conditions and operating conditions should be related, but not identical in risk. A new worker can learn through controlled variations and realistic simulations rather than unsupervised exposure to hazards. The manager also has to manage work design: more practice will not compensate for impossible staffing, conflicting instructions, unavailable equipment or missing authority to decide.

• Execute. Apply the normal process without an extra prompt.
• Recognise. Identify exceptions, uncertainty and stop criteria.
• Escalate. Use the correct role and channel before risk grows.

This matters in cold-chain handling, transport safety and hotel operations. The point is not to reveal employer-specific controls; it is that a trainee who safely stops an unauthorised action may have shown better judgement than one who completes the task quickly by guessing. A balanced assessment never rewards speed at the expense of compliance or service quality.

Our reading

Competence includes knowing when to pause, not only how to move quickly.

Research lens: Keith and Frese (2008), Grossman and Salas (2011); the task-risk application is editorial.

Sources: Nina Keith & Michael Frese, Journal of Applied Psychology 93(1), 2008; Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), 2011

Layer 2 · The conceptual and scientific breakdown

What learning from errors can mean

Error-management training has research support, but it must not be confused with permission for uncontrolled errors in live operations.

24 studies, with 2,183 participants, in Keith and Frese's 2008 meta-analysis of error-management training

Standardised mean effects reported in the meta-analysis, scaled to d = 0.80: Overall mean d = 0.44, Post-training transfer d = 0.56, Adaptive transfer d = 0.80.

These estimates refer to different outcome subsets and comparisons; they are not percentages, success rates or a comparison of three programmes. The researchers found a positive mean effect across the included studies and larger effects for post-training and adaptive-transfer outcomes in moderator analyses. The evidence concerns active exploration and explicit learning from mistakes in designed training activities. In a hotel or loading operation, the responsible translation is supervised practice with realistic but safe variations. Hazardous live procedures cannot be used as an informal experiment, and the meta-analysis guarantees neither improvement in a particular shift nor an equal benefit for every novice.

Our reading

An error can be a teaching signal in a controlled exercise. In live operations, safety and authorisation still govern action.

Keith and Frese, Journal of Applied Psychology 93(1), 59–69 (2008). The three values are pooled estimates from the paper, drawn to scale.

Source: Nina Keith & Michael Frese, Journal of Applied Psychology 93(1), 2008

Layer 3 · The operating protocol for action

Rehearse variation without adding risk

A progressive practice ladder that challenges judgement in conditions approved for training.

• Define the routine task. Its observable acceptance criteria and its safety boundaries.
• Run a normal rehearsal. Completion with minimal prompting in a permitted training environment.
• Change one condition. A simulated exception, interruption or ambiguous status.
• Ask for the reason. What changed, what does the standard require and what remains unknown?
• Practise the safe boundary. Stopping, seeking help or escalating instead of inventing a shortcut.
• Review both person and system. Was the miss a learning gap, an unclear instruction or a condition the operating design must repair?

Document the practice task and the kind of assistance provided. Do not score a simulated scenario as a measured workplace outcome. Repeated rehearsals serve learning; readiness for an actual role still requires employer-approved checks. Where a person fails to demonstrate a high-risk task, continued supervision is a control, not a punishment.

Our reading

Make mistakes safe to examine during learning; do not make unsafe exceptions normal during service.

Editorial practice protocol informed by Keith and Frese (2008), Salas et al. (2012) and local-authority safeguards. The site's 5 Whys worksheet serves the last step.

Sources: Nina Keith & Michael Frese, Journal of Applied Psychology 93(1), 2008; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

Chapter V: From training room to real work

Layer 1 · The living scene

The course that never reached the shift

The programme can be well delivered while the workplace quietly prevents people from using what they learned.

Hypothetical training-transfer scene, not a measured programme

A new employee returns from an approved training session. The exercise went well: the person can explain the standard and demonstrates the expected sequence in practice. Back on the shift, the colleague meets a different pattern of work. The person who could answer questions is assigned elsewhere, and the task is immediately handled through a long-established local habit. Several weeks later, the team still uses the old pattern. A superficial evaluation could conclude that the trainee forgot the lesson. Other explanations are possible: no opportunity to practise the target behaviour, no protected time to learn, an outdated system, or a leader who never clarified that the new standard had actually been adopted. A class cannot transfer itself into an unsupportive process.

This scene is constructed to isolate the environmental component of transfer. It does not assert that the author introduced a programme that failed or that a particular hotel or distribution partner discouraged new methods. The research question is whether the learning outcome was measured separately from workplace behaviour.

Our reading

If the environment prevents a learned behaviour from being used, repeated classroom instruction may not solve the problem.

Editorial illustration linked to Baldwin and Ford (1988) and Grossman and Salas (2011). The essay named below follows a new system into the work.

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), 2011

More in the essay: When a new system arrives

Layer 2 · The conceptual and scientific breakdown

The workplace is part of the course

The supervisor, the tools and the opportunity to practise determine whether learning can become everyday performance.

Transfer begins before the session and continues after it. A manager who sponsors training has to explain why the competence matters, reserve relevant practice opportunities, provide feedback and remove contradictions between the new instruction and the workflow. Training professionals can design a good learning activity, but they cannot independently create the conditions in which the shift will use that skill. Grossman and Salas (2011) identify support, opportunity to perform and follow-up among the factors consistently associated with transfer. Burke and Hutchins (2007) treat the work environment as a distinct source of transfer conditions. Salas and colleagues (2012) organise training across what happens before, during and after a programme. Sitzmann and Weinhardt (2018) describe training engagement across people, workgroups and organisations, and across time.

• Before. Define need, support and criteria.
• During. Practise with feedback.
• After. Observe use, reinforce and adapt.

At the level of one team this is concrete. Was the new employee scheduled to use the learned task? Did a qualified person review the first independent attempts? Was uncertainty recorded and resolved? Do the SOP versions used in class and in the workplace match? If the answers are not known, the evaluation is incomplete.

Our reading

Transfer is a joint property of the training and the conditions in which the trained person is expected to work.

Burke and Hutchins (2007); Grossman and Salas (2011); Salas et al. (2012); Sitzmann and Weinhardt (2018).

Sources: Lisa A. Burke & Holly M. Hutchins, Human Resource Development Review 6(3), 2007; Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), 2011; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012; Traci Sitzmann & Justin M. Weinhardt, Journal of Management 44(2), 2018

Layer 2 · The conceptual and scientific breakdown

Four outcomes, four different questions

Training-effectiveness research does not license managers to treat satisfaction scores as evidence of customer results.

Published sample-weighted mean effect sizes (d), scaled to d = 0.63: Reaction (k = 15) 0.60, Learning (k = 234) 0.63, Behaviour (k = 122) 0.62, Results (k = 26) 0.62.

Arthur, Bennett, Edens and Bell (2003) meta-analysed studies of organisational training and reported separate sample-weighted effect sizes for reaction, learning, behaviour and results. The pooled estimates were positive, but the criteria were assessed in different sets of studies: these are distinct estimates, not a percentage trained or one group followed through four stages. The authors also found that training method, task characteristics and the choice of outcome criterion were related to measured effectiveness. A pleased trainee may have learned little; a skilled trainee may have no chance to perform; a change in individual behaviour may be too small or too loosely connected to a wider customer outcome. The four measures ask different questions and cannot be substituted for one another, nor do they qualify anyone for a regulated or safety-critical task. The contribution for a practical manager is measurement humility: define the outcome first, then choose the evidence that answers that question.

Our reading

A training survey cannot certify workplace behaviour, and a classroom test cannot certify the organisation's results.

Arthur et al., Journal of Applied Psychology 88(2), 234–245 (2003). The four values and their k are the paper's pooled estimates, drawn to scale.

Source: Winfred Arthur Jr., Winston Bennett Jr., Pamela S. Edens & Suzanne T. Bell, Journal of Applied Psychology 88(2), 2003

Layer 3 · The operating protocol for action

Make transfer visible after class

A short set of observations can separate delivery, learning and use without turning a trainee into a surveillance target.

• Before training. Task, approved standard, risk level and expected observable behaviour.
• At completion. Explanation and guided demonstration, with the support required recorded.
• At first application. An authorised real or simulated task, with where help was needed classified.
• After an agreed interval. The same ability under relevant conditions, not only memory of the guide.
• Review barriers. Equipment, time, workload, unclear SOPs or supervisor practices.
• Report at the correct level. Capability gaps and process defects, no individual performance data published.

An interval of a week or a month is a local choice, not a universal requirement. Keep the same task definition, track quality and unintended consequences (faster completion that creates downstream rework is not transfer), and ask the people who receive the work.

Our reading

The key outcome is not that a session took place. It is that a useful behaviour appeared when the work required it.

Editorial method synthesised from Arthur et al. (2003), Grossman and Salas (2011) and Salas et al. (2012).

Sources: Winfred Arthur Jr., Winston Bennett Jr., Pamela S. Edens & Suzanne T. Bell, Journal of Applied Psychology 88(2), 2003; Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), 2011; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

Chapter VI: Proving readiness

Layer 1 · The living scene

Two certificates, different evidence

The same training completion status can conceal two very different levels of independence.

Constructed assessment scene, not a real personnel decision

Two new colleagues finish the same induction. Both attend all sessions, acknowledge the procedures and complete the supervised demonstration. A manager could mark both as ready. Instead, the assessor asks each person to run an approved task and explain what would make the next action inappropriate. One trainee describes the boundary and demonstrates how to seek the correct help. The other completes the normal task but needs prompting when the variation appears. This comparison is deliberately fictional. It does not establish that one worker has greater potential, that the other is unmotivated or that an organisation used such a test. It shows why a readiness decision should describe the task and the evidence. The second colleague may need a clearer example, more rehearsal or more protected practice. The response should be developmental and proportionate to risk, not a public label attached to a person.

The authorised supervisor decides the scope of independent work, and some roles require formal approvals outside any local training checklist. This volume creates no credentials and reproduces no confidential operating steps. Its concern is the quality of evidence behind a decision that training is finished.

Our reading

A fair readiness decision describes demonstrated capability and the support still required, not a judgement of personal worth.

Constructed illustration informed by Baldwin and Ford (1988), Arthur et al. (2003) and Salas et al. (2012). The essay named below tests an SOP in the shift, not in the folder.

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Winfred Arthur Jr., Winston Bennett Jr., Pamela S. Edens & Suzanne T. Bell, Journal of Applied Psychology 88(2), 2003; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

More in the essay: A good SOP is not a document

Layer 2 · The conceptual and scientific breakdown

Five kinds of readiness evidence

A progression of observable evidence, without pretending that every role has one universal qualification pathway.

• Exposure. The employee was shown the current approved process; attendance and version recorded.
• Understanding. The employee can explain the purpose, expected result and relevant limits.
• Supported practice. The employee performs in a permitted setting with assistance made explicit.
• Independent evidence. A bounded task without inappropriate prompting, with exceptions recognised.
• Maintained transfer. Later observation finds the capability usable in relevant working conditions.

These are editorial evidence categories, not a validated psychometric scale. A role may require parallel assessments, formal external certification or mandatory supervision regardless of one demonstration. A correct stop-and-escalate decision can count positively where that is the approved standard; performing a normal sequence once does not establish unlimited authority for unusual cases. A readiness record names the assessed process, SOP version, task risk, date, assessor, observed evidence, remaining gaps and next review. Avoid aggregate pass rates across incomparable roles, and never use the card to make sensitive employee data public.

Our reading

Readiness is a bounded claim supported by observed work, not a permanent personal label.

Original operating framework informed by the transfer and evaluation literature (Baldwin and Ford 1988; Arthur et al. 2003; Blume et al. 2010; Salas et al. 2012); not a validated readiness score or employer certification.

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Winfred Arthur Jr., Winston Bennett Jr., Pamela S. Edens & Suzanne T. Bell, Journal of Applied Psychology 88(2), 2003; Brian D. Blume, J. Kevin Ford, Timothy T. Baldwin & Jason L. Huang, Journal of Management 36(4), 2010; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

Layer 3 · The operating protocol for action

A controlled readiness pilot

Test the assessment design before scaling it: safe, fair and useful for the receiving team.

• Select one task family. A common, authorised task with observable quality criteria.
• Confirm the rule owner. The responsible department validates the instruction and stop conditions.
• Pilot with supervision. Task-level observations, no uncontrolled risk for trainees.
• Measure assistance separately. Complete, accurate and supported performance, not simply speed.
• Check for transfer. A repeat observation after an agreed interval, with changed conditions noted.
• Review the system as well. Obstacles, trainer prompts, SOP clarity and downstream rework.

Ten anonymised observations are an editorial illustration, not a validated sample size. A local owner agrees in advance what counts as success; the trainee needs a safe route to ask, and the programme a way to revise confusing materials. A workshop that finds an unclear procedure has found information about the process, not a person to blame. If the capability is not demonstrated, supervision and more practice are concrete actions; a certificate without evidence for its scope is not.

Our reading

Do not certify the course. Verify the task, the conditions and the limits of the competence claim.

Editorial pilot after Baldwin and Ford (1988) and Salas et al. (2012); not a ready-made regulatory scheme.

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

Field tool

The Operational Readiness Card

One role, one bounded task. Complete only with anonymised or properly authorised records, and write “not observed” where evidence is missing.

1. Task and standard which approved task, SOP version and expected result are being assessed?
2. Demonstration what did the trainee perform without prompting, and what help was given?
3. Exception recognition what boundary or stop-and-escalate condition did the trainee identify?
4. Quality and transfer what evidence shows the task is accurate and usable downstream?
5. Support and ownership what further practice is required, and who can authorise the next step?
6. Follow-up when is the next observation, and how will the result return to the trainee?

A proposed educational aid, not a validated test or employer-issued certification. A live assessment follows the employer's training, privacy, safety and approval requirements; never store customer data, guest details, protected personnel records or credentials in an informal tool, and never read speed alone as competence.

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012

Open the tool: 5 Whys

Closing essay

The day the trainer leaves

Every training programme eventually arrives at a moment that the classroom cannot control.

An experienced colleague finishes a demonstration and turns back to the ordinary work of the day. The new employee now faces the job without the same prompts, patience and predictable sequence. In one sense the programme has ended. In another, the important part is only beginning: whether knowledge can travel from the protected practice setting into decisions made under real conditions. Managers need training records. They also need competent people, responsible limits and a way to repair the process when someone cannot perform what the organisation demands. The mistake is not counting completion. It is treating that count as the entire result. A signature, a quiz or a successful guided task answers a narrow question. Neither the person nor the organisation benefits when that narrow answer is inflated into a claim of total readiness.

The research offers discipline rather than a magic technique. Behaviour-modelling studies say that watching and learning are not identical to later job performance. Transfer reviews point to practice opportunity, workplace support and follow-up. Error-management research suggests that safely exploring mistakes in training can help prepare for unfamiliar situations. None of this licenses managers to weaken real safety boundaries or to claim that every novice should perform independently after the same number of hours. The author's experience spans new-driver ramp-up, last-mile coordination and the search for reliable hotel SOP guidance while learning a new role. Those experiences make the question concrete, but this volume does not measure failures or successes in any named workplace. Its practical value lies in asking a more demanding question before writing “training completed”: what did the learner actually demonstrate, what remains uncertain, and who will help if the real situation differs from rehearsal? An organisation earns trust when it can answer that question without blaming the beginner for an incomplete training system. The trainer leaves. The work continues. Readiness is the evidence that the employee is not being sent into that gap alone.

Our reading

Training is not complete when someone has watched the work. It is complete for a defined task when performance, limits and support are demonstrated.

Editorial conclusion. No failed programme, unready employee or unsafe practice is attributed to any named employer.

Method

Evidence before the readiness claim

An evidence-informed editorial synthesis, not original field research: documented categories of last-mile ramp-up and hotel training work, labelled constructed scenes and proposed operating protocols. Every numeric sample and effect refers to published research, never to the author's workplaces.

• Reviews. Baldwin and Ford (1988), Burke and Hutchins (2007), Aguinis and Kraiger (2009), Grossman and Salas (2011), Salas et al. (2012) and Sitzmann and Weinhardt (2018) organise the literature; they measure no local readiness.
• Meta-analyses. Arthur et al. (2003): d = 0.60, 0.63, 0.62, 0.62 for reaction, learning, behaviour, results (k = 15, 234, 122, 26). Taylor et al. (2005): 117 studies. Keith and Frese (2008): 24 studies, 2,183 participants, d = 0.44, 0.56, 0.80. Blume et al. (2010): 89 studies.
• Charts. Both charts draw published pooled estimates to scale; each value comes from a different outcome subset, so none is a readiness rate or a comparison of programmes.
• Documented context. One early hotel training night limited by a system update, and the author's Front Desk Control reference; no before-and-after skill level and no measured effect of the tool.

The Operational Readiness Card, the practice ladder and the readiness pilot are editorial proposals for local assessment; they are not validated tests or employer-issued certifications and must not bypass training authority. No personnel, guest, payment or confidential system details are published. Editorial standard: a review, a meta-analysis, a documented context and a proposal are named as such, and the sentence stops claiming where the evidence stops.

Every source is cited at its DOI on the next page; the ten records, their journals, volumes and years were checked against the publishers' metadata.

Sources: Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), 1988; Winfred Arthur Jr., Winston Bennett Jr., Pamela S. Edens & Suzanne T. Bell, Journal of Applied Psychology 88(2), 2003; Paul J. Taylor, Darlene F. Russ-Eft & Daniel W. L. Chan, Journal of Applied Psychology 90(4), 2005; Lisa A. Burke & Holly M. Hutchins, Human Resource Development Review 6(3), 2007; Nina Keith & Michael Frese, Journal of Applied Psychology 93(1), 2008; Herman Aguinis & Kurt Kraiger, Annual Review of Psychology 60, 2009; Brian D. Blume, J. Kevin Ford, Timothy T. Baldwin & Jason L. Huang, Journal of Management 36(4), 2010; Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), 2011; Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), 2012; Traci Sitzmann & Justin M. Weinhardt, Journal of Management 44(2), 2018

Sources and method

Where the claims come from

The figures in this volume come from the sources below, each with its year. The scenes are constructed; the protocols are proposals; what is research is named as research.

• Timothy T. Baldwin & J. Kevin Ford, Personnel Psychology 41(1), “Transfer of Training: A Review and Directions for Future Research”, 1988. https://doi.org/10.1111/j.1744-6570.1988.tb00632.x
• Winfred Arthur Jr., Winston Bennett Jr., Pamela S. Edens & Suzanne T. Bell, Journal of Applied Psychology 88(2), “Effectiveness of Training in Organizations: A Meta-Analysis of Design and Evaluation Features”, 2003. https://doi.org/10.1037/0021-9010.88.2.234
• Paul J. Taylor, Darlene F. Russ-Eft & Daniel W. L. Chan, Journal of Applied Psychology 90(4), “A Meta-Analytic Review of Behavior Modeling Training”, 2005. https://doi.org/10.1037/0021-9010.90.4.692
• Lisa A. Burke & Holly M. Hutchins, Human Resource Development Review 6(3), “Training Transfer: An Integrative Literature Review”, 2007. https://doi.org/10.1177/1534484307303035
• Nina Keith & Michael Frese, Journal of Applied Psychology 93(1), “Effectiveness of Error Management Training: A Meta-Analysis”, 2008. https://doi.org/10.1037/0021-9010.93.1.59
• Herman Aguinis & Kurt Kraiger, Annual Review of Psychology 60, “Benefits of Training and Development for Individuals and Teams, Organizations, and Society”, 2009. https://doi.org/10.1146/annurev.psych.60.110707.163505
• Brian D. Blume, J. Kevin Ford, Timothy T. Baldwin & Jason L. Huang, Journal of Management 36(4), “Transfer of Training: A Meta-Analytic Review”, 2010. https://doi.org/10.1177/0149206309352880
• Rebecca Grossman & Eduardo Salas, International Journal of Training and Development 15(2), “The Transfer of Training: What Really Matters”, 2011. https://doi.org/10.1111/j.1468-2419.2011.00373.x
• Eduardo Salas, Scott I. Tannenbaum, Kurt Kraiger & Kimberly A. Smith-Jentsch, Psychological Science in the Public Interest 13(2), “The Science of Training and Development in Organizations: What Matters in Practice”, 2012. https://doi.org/10.1177/1529100612436661
• Traci Sitzmann & Justin M. Weinhardt, Journal of Management 44(2), “Training Engagement Theory: A Multilevel Perspective on the Effectiveness of Work-Related Training”, 2018. https://doi.org/10.1177/0149206315574596

Editorial method

Each figure was checked for its year, its publisher and what exactly it measures, at the full text where it could be opened. Where the publisher's page could not be opened, the figure was checked against independent summaries and is marked “via”. The editors' interpretation is marked “Our reading”. No figure of the author's own workplaces is claimed.