Job Analysis, Evaluation and Design

An auto-parts plant in Pune needs twenty new machine operators. Before it can advertise, test candidates or fix their pay, someone has to know exactly what an operator does all day, and what it takes to do it well.

Job analysis yields four kinds of information:

• Purpose and nature: why the job exists and what it involves.
• Organisational fit: its department, reporting line and place in the hierarchy.
• Human profile: the skills, attributes and behaviour it needs.
• Relationships and context: whom the jobholder deals with, what they depend on, the tools they use and the demands they face.

Back to the plant. As an illustration, an analyst watches operators for a week, interviews two supervisors, and collects accounts of the best and worst shifts. She finds that most of an operator's day goes on setting machines, and that the costliest mistakes happen at shift change. That one finding shapes the job advertisement, the selection test and the training plan.

Methods. No single method sees everything, so most analyses combine two or more.

Textbooks also list questionnaires and work diaries, in which jobholders record what they do over a period.

Strategic uses. Job analysis supports work simplification (breaking complex jobs into manageable tasks), sets measurable performance standards, and supports recruitment, selection, training and appraisal.

Don't confuse job analysis with job evaluation. Analysis describes what a job is and what it needs. Evaluation judges what the job is worth compared with other jobs. Analysis comes first and feeds evaluation.

Where it breaks. Jobs change faster than documents. In fast-moving roles, a careful analysis can be out of date within a year. Each method also has a blind spot, as the table shows.

For exams. Expect to match a method to its description. The critical incident technique is linked with John Flanagan (1954); time study with F. W. Taylor; motion study with Frank and Lillian Gilbreth.

A limit of observation: people may behave differently when watched (the Hawthorne effect).

A public-sector bank has 1,200 clerks across 300 branches. One analyst cannot sit beside every clerk. She needs ways to collect facts from many people, and ways to check what they tell her.

Observation, work method analysis and the critical incident technique are covered in the first block of this chapter. This block adds the methods that rely on the jobholder's own account.

Work method analysis in two levels. Time and motion study times each task and the movements it needs, to find the best sequence. Micro-motion analysis goes finer. It films a task and studies the movements frame by frame, as the Gilbreths did. Both suit repetitive manual jobs.

Back to the bank. As an illustration, the analyst sends a questionnaire to all 1,200 clerks. She interviews 20 of them with a fixed set of questions, and interviews their supervisors too. Ten clerks at busy branches keep a diary for a fortnight. Suppose the questionnaires say that helping customers fill in forms takes little time, but the diaries show about an hour a day. A gap between two methods is a question, not an answer. She asks the supervisors, and finds that the forms were redesigned last year.

Don't confuse the work diary with the critical incident technique. The diary records the ordinary day, task by task. Critical incidents record only the extremes, the exceptionally good or poor episodes. A diary shows how time is spent; incidents show what separates good work from poor.

Where it breaks. All four methods depend on what people say or remember. Jobholders who think the study will change their pay or workload may exaggerate or understate. The remedy is to use two methods and compare, and to check claims with supervisors.

In practice. Let the job decide the method.

• Many people in the same job: questionnaire.
• A few complex or unique jobs: interviews.
• Irregular work with odd tasks: diary.
• Routine physical work: observation or work method analysis.

For exams. Know the interview types (structured; unstructured, also called open-ended) and that work method analysis includes time and motion study and micro-motion analysis. Standard questionnaires such as the PAQ and CMQ, and Sidney Fine's Functional Job Analysis, are described in the competency mapping chapter.

Watch it, ask about it, or let the jobholder write it down.

A bank branch advertises for "a smart, hardworking officer". Two hundred people apply, and most are wrong for the job. A clear job description and job specification would have filtered them out before anyone read a CV.

Both are written from the findings of job analysis. One looks at the work; the other looks at the worker.

An example. Here is an illustration for a counter supervisor at a public-sector bank branch.

How they are used. The description becomes the job advertisement, the induction plan and the base for appraisal goals. The specification becomes the screening checklist and the interview plan.

Don't confuse the two. A common slip is to put "three years' experience" in the description. Anything about the person goes in the specification; anything about the work goes in the description.

Where it breaks. A rigid description invites "not my job" thinking, so many firms write broader role profiles. A specification can be inflated, such as a postgraduate degree for data entry. That shrinks the pool of applicants and can shut out people who could do the job well.

In practice. Test each line of the specification against the description: which duty needs this? If no duty needs it, drop it.

The job description says what the job involves; the job specification says who can do it.

KSAO. Many textbooks summarise a specification as the jobholder's knowledge, skills, abilities and other characteristics (KSAOs), such as temperament and values. Some notes write the last letter as C, for characteristics. The idea is the same: list what the person must bring, not what the work involves.

At a milk cooperative, the lab chemist earns less than the truck dispatcher and asks why. Without a method for comparing jobs, the only answer is "that is how it has always been", and that answer breeds resentment.

Objectives. Internal equity, so that jobs of similar worth get fair pay; a standard pay structure; fairness, with less favouritism and wage inequality; and benchmarking against industry pay.

The process runs in four steps:

Study each job, write it up, rate it with one of the methods below, then arrange the jobs in order and group them into grades. Pay ranges are attached to the grades afterwards.

The points system, worked. A compensable factor is a feature of a job that the organisation is willing to pay for. Say the cooperative uses four, each with five levels. The maximum points show each factor's weight.

A panel rates each job on each factor:

• Lab chemist: skill level 5 (400), responsibility 4 (240), effort 2 (80), conditions 3 (60). Total 780.
• Truck dispatcher: skill 3 (240), responsibility 3 (180), effort 3 (120), conditions 2 (40). Total 580.
• Packing-line worker: skill 1 (80), responsibility 1 (60), effort 4 (160), conditions 4 (80). Total 380.

The chemist's job ranks highest. If grades are set at 700 points and above, 500 to 699, and below 500, the three jobs fall into three different grades. The chemist's complaint was justified.

Factor comparison works differently. It starts with benchmark jobs whose pay is accepted as fair, splits that pay across the factors, and then slots other jobs in factor by factor.

Don't confuse job evaluation with performance appraisal. Evaluation rates the job; appraisal rates how well a person does it. Two operators in the same job share one job value but can get different appraisal ratings.

Where it breaks. Choosing the factors and their weights is a judgement, so bias can enter dressed up as numbers. Evaluation sets the internal order, but the job market can pull against it when a skill is scarce. And it needs redoing as jobs change.

For exams. Know which methods are quantitative, the order of the process, and the four objectives: internal equity, a standard pay structure, fairness and benchmarking.

Non-quantitative methods compare whole jobs; quantitative ones score factors.

A dairy cooperative must evaluate six jobs this month and two hundred next year. The method that works for the first lot will not work for the second, and the committee needs to know why.

Ranking. A panel puts jobs in order of overall worth. In paired comparison it compares every job with every other and counts wins. Six jobs need 15 comparisons (6 × 5 ÷ 2). Two hundred jobs would need 19,900 (200 × 199 ÷ 2). That is why ranking breaks down as jobs multiply.

Classification. The committee writes the grades first. Each grade has a description covering factors such as education, independent working, knowledge, skills and decision-making. Then every job is matched to the grade whose description fits best, and pay ranges attach to the grades.

Factor comparison, worked. The panel follows five steps.

1. Choose key jobs whose pay is agreed to be fair by management and workers.
2. Choose the factors, here skill, responsibility and effort.
3. Rank the key jobs on each factor.
4. Divide each key job's pay among the factors, in line with those ranks.
5. Rate a new job factor by factor against the key jobs, and add the rupee values.

The numbers are an illustration. Now rate a stores clerk. The panel places the clerk's skill between the packing worker and the dispatcher (₹7,000), responsibility in the same gap but nearer the packing worker (₹6,000), and effort just below the packing worker's and level with the dispatcher's (₹9,000). The clerk's job is worth 7,000 + 6,000 + 9,000 = ₹22,000 a month.

Don't confuse ranking with classification. Ranking compares jobs with one another. Classification compares each job with a fixed written description. Also, the points system scores in points and factor comparison in rupees. Some books call factor comparison a hybrid, because it ranks jobs and then prices them; most classify it as analytical.

Where it breaks. Factor comparison inherits any unfairness in the key jobs' pay. If a key job is underpaid because of old bias, every job rated against it is underpaid too. The analytical methods do not remove judgement. They split it into smaller decisions, which are easier to see and to challenge.

For exams. Textbooks list more objectives than the four given earlier. They add:

• complete and impersonal descriptions of every job;
• one standard procedure for judging worth;
• like pay for like work;
• a fair basis for promotion and transfer;
• facts for comparing pay with similar jobs outside.

Evaluation fixes the internal order. Comparing it with outside pay is a separate step, done with pay surveys.

Ranking: job against job. Classification: job against grade. Points: factor scores. Factor comparison: factor rupees.

A garment unit splits shirt-making into thirty tiny steps. Output rises at first, but so do absences: people who stitch one collar seam all day get bored and leave. How tasks are bundled into jobs matters.

An example. In a bank's loan section, a clerk only types application data into the system. That is a simplified job.

• Rotation: the clerk spends three months each in deposits, loans and foreign exchange.
• Enlargement: the clerk now enters the data, checks the documents and files them. There are more tasks, all at the same level.
• Enrichment: the clerk handles small loans from start to finish and can approve them up to a set limit. The job now carries responsibility.

Don't confuse enlargement with enrichment. Enlargement is horizontal: more of the same kind of work. Enrichment is vertical: more control over the work. Frederick Herzberg warned that piling on more routine tasks ("horizontal job loading") does not motivate.

Where it breaks. Enrichment does not suit everyone; some people prefer clear, routine work. It also needs training, and people given more responsibility often expect more pay. Simplification still suits high-volume, standardised work done by new staff.

In practice. Match the method to the problem. Boredom calls for rotation or enlargement. Capable staff with little ownership call for enrichment. Frequent errors in high-volume work may call for simplification.

Simplification: fewer tasks. Rotation: different tasks. Enlargement: more tasks. Enrichment: more authority.