Indirect Tax and GST Compliance Assessment for Finance, Accounts and Billing TeamsThe total was right. The heads were wrong. Which one would you have stopped?
Forty exercises on applying goods-and-services-tax rules to the documents a finance desk actually handles. Ten print a rule and six lines of an invoice, a register or a reconciliation, and you mark the lines that would stop it; the number of defects varies, so counting is no help. Every rule is stated in the exercise, so no current rate, threshold or due date is the answer to anything. The report counts the two ways you can be wrong apart and never averages them.
A report that prices the two errors differently, because a filing does
The Indirect Tax and GST Compliance Assessment for Finance, Accounts and Billing Teams is a thirty-five-minute knowledge check of whether a person can apply a stated goods-and-services-tax rule to an invoice, a credit claim or a reconciliation and tell the line that would stop it from the line that is fine, reported as a two-error panel.
Every exercise states the rule it uses in one plain sentence and then asks you to apply it. That is a deliberate design, not a concession: rates, thresholds and due dates change, and a test keyed to a remembered rate measures memory of last year's notification, not judgement. What does not change is the structure — place of supply, forward against reverse charge, what credit is allowed and what is blocked, the difference between a tax invoice, a bill of supply, a credit note and a debit note, exempt against zero-rated, and what a mismatch between the books and the filed return actually means. Where a figure is needed, it is given.
Ten of the forty exercises are document sweeps: six lines of a draft invoice, a credit register or a month-end reconciliation, with a hidden class on every line and between one and four defects per document. The defects that get through on these are the quiet ones — a total that is right split into the wrong tax heads, a listed service carrying the supplier's tax, a credit note nobody reported — and the lines most often held by mistake are the rule working as written. Ten more exercises print one situation and three specific defect claims plus a nothing-would-stop-it option, so the report can tell a person who senses something is wrong from a person who can say what it is.
Scoring is signal detection. Two numbers are reported, not one: how well you tell the two classes apart, and where you set your bar for holding a line. Holding everything and passing everything both land at zero on the first and are told apart only by the second, so neither habit can pass for reading. A respondent answering the way people typically answer these exercises — modelled from a declared prior on every line — sits above that zero, and is drawn on the same axis as you, so the page shows where you stand against a typical desk and not only against a coin.
The report is a two-error panel. A defect let through and a comment spent on a line that was fine are counted apart, each with its own cost sentence in a finance desk's own setting, a worked example from your own answers, and its own fix. The declared cost matrix — a let-through costs twice a false alarm — is printed above the two columns because it is a value judgement, and the expensive column is drawn heavier and says the ratio in words. Beneath them, one line says which of the two you do more and what that habit costs at the printed prices. Every figure carries its 68 and 95 per cent band, drawn and written in words; an area with too few answers or too little reliability carries a placement instead of a number, and the refusal is printed where the number would have been.
It is not tax advice, not a qualification, and not affiliated with any tax authority, and the report says so before it says anything else. No statute section, form name, notification or filing portal is the key to any exercise, and no real business, supplier or officer is named anywhere. It measures one thing: given the rule, can you see which line stops the document.
What you walk away with
Where a supply of goods or services is placed, when one integrated tax applies and when the tax splits, and which listed supplies the recipient pays for directly.
The conditions a credit must meet before it is claimed, what happens at 180 days unpaid, credit on goods arriving in lots, common credit shared with exempt supplies, and the blocked categories.
Which document each situation takes, what a tax invoice must carry, who may issue the note that changes a supplier's tax, and how each kind of invoice error is corrected.
Why exempt and zero-rated look alike on the invoice and differ underneath, what goes into the value and what comes out, and which rate a bundle takes.
What a difference between the register, the return and the auto-drawn statement actually means — timing, a wrong number, a missing report — and which of them needs action.
How often you let a defect through and how often you hold a line that was fine, each with its own cost, its own fix and a worked example from your own answers, plus whether you can name what you sensed.
Inside your report
Illustrative sample — your report is generated from your own responses.
typical about 13.7
typical about 7.7
| Area | Answered | Omega | Figure or printed refusal |
|---|---|---|---|
| Place of supply and who pays the tax | 8 of 8 | 0.716 | +26 · Above the typical respondent · 68% 13 to 39 · 95% 1 to 51 |
| Input tax credit and blocked credits | 8 of 8 | 0.716 | +4 · Near the typical respondent · 68% -9 to 17 · 95% -21 to 29 |
| Tax invoice, bill of supply, credit and debit notes | 7 of 8 | 0.689 | Near the typical respondent · figure withheld: 7 answered is under the 8 a figure needs |
Built for
- Accounts payable and receivable teams who check supplier invoices and raise customer invoices under a goods-and-services tax
- Billing staff and tax executives who prepare the monthly return and the reconciliation behind it
- Finance shared-services teams onboarding new joiners onto indirect-tax work, and the managers who sign off their queues
- Small-business owners and their bookkeepers who handle their own invoicing and credit claims
Find out which of the two errors you make more, and what it costs at the printed ratio
40 exercises across six formats · about 35 minutes · every rule stated in the exercise, two error directions counted apart, and a report that is not tax advice and says so first.
₹699 (incl. GST) · assessment and full report, nothing further to pay
Frequently asked questions
Whether you can apply a stated goods-and-services-tax rule to a document, a claim or a reconciliation and tell the line that would stop it from the line that is fine, across five areas: place of supply and who pays the tax, input tax credit and blocked credits, the tax invoice, bill of supply, credit note and debit note, exempt and zero-rated supplies and the value of a supply, and books-to-return reconciliation. It does not measure whether you know any current rate, threshold, due date or form.
No. Every exercise states the rule it uses in one plain sentence, and where a figure is needed it is given. Rates and dates change; the structure does not. The construct is applying a stated rule correctly, not recalling a current one. Nothing in the report is tax advice, and it is not affiliated with or endorsed by any tax authority.
Because they cost different things and need opposite fixes. A defect let through reaches the return and comes back as interest, penalty exposure and a reversal. A line held that was fine is an unpaid invoice, a supplier chasing and a colleague re-reading a correct document. The report counts each, prices each against a matrix that is printed on the page, shows you one of your own answers in each column, and never averages them.
No. Holding everything and passing everything both land at zero on the discrimination figure and are told apart only by where you set your bar. Both habits are priced beside you at the printed costs, and neither beats a respondent answering the way people typically answer these exercises. The only way through is to read the rule and apply it to the line.
No. It is a measurement of how you applied forty stated rules in one sitting of about thirty-five minutes. It is not a qualification, it confers nothing, it is not tax, legal or accounting advice, it is not a substitute for a professional, and no score here means anything to any tax authority or professional body. The report prints that before it prints any number.
Each one takes a single capability, puts you inside the situations where it is actually tested, and scores your choices against published evidence — with a report designed for that capability alone, not a template. They span hiring, compliance, education, operations and personal skill.
Browse the catalogue →Methodology: Forty original exercises across six formats: ten document sweeps, each printing the rule that governs it and six lines of an invoice, a register or a reconciliation, of which between one and four are defects and the rest are ordinary, and the respondent marks the lines that would stop the document; ten naming exercises, each printing a rule and one situation with three specific defect claims and a nothing-would-stop-it option, of which seven carry a defect and three are clean; seven true-or-false claims that apply a stated rule; five match-the-following exercises; four ordering exercises; and four numeric estimation exercises on rate and value arithmetic with every figure given in the exercise. One response instruction is declared for the whole instrument and it is a KNOWLEDGE instruction: given this rule, what would stop this document, what is wrong with this line, what is the tax on this value; never what the respondent would do. Construct statement: this instrument measures whether a person can apply a stated goods-and-services-tax rule to a document, a claim or a reconciliation and tell the line that would stop it from the line that is ordinary, across five areas: place of supply and who pays the tax, input tax credit and blocked credits, the tax invoice, bill of supply, credit note and debit note, exempt and zero-rated supplies and the value of a supply, and books-to-return reconciliation. It does not measure whether the respondent knows any current rate, threshold, due date or form, it does not measure honesty, diligence or behaviour under filing pressure, it is not tax, legal or accounting advice, it is not a qualification and confers none, and it is not affiliated with, endorsed by or derived from any tax authority, tax board, professional body or filing platform. Rules change; every exercise states the rule it uses, and the construct is applying a stated rule correctly, not recalling a current one. Where a figure is needed it is given in the exercise. Scoring is signal detection. Across the ten sweeps and the ten naming exercises every line or situation carries a hidden class, defect or ordinary, and the number of defects varies from one to four per sweep so that counting to a fixed number is not a strategy; every sweep carries at least one defect and at least one ordinary line, so both errors can occur on it. The hit condition is the classic flag-or-leave call: marking a defect line, or choosing any defect claim on a naming exercise, is a flag. Discrimination d-prime is z of the hit rate minus z of the false-alarm rate and the criterion c is minus one half of their sum, both with the log-linear correction of one half per cell so that an extreme rate stays finite. Both are reported, because flagging everything and flagging nothing both land on a d-prime of exactly zero as arithmetic and are told apart only by c: a flagger and a filterer at the same d-prime are different people. The hit condition was chosen after computing what a respondent drawing from the declared priors scores on it: on this file, 33 defect calls and 37 ordinary calls, that respondent has a hit rate of 0.586, a false-alarm rate of 0.209, a d-prime of 1.03 and a criterion of 0.30, above the zero that both fixed habits land on, so no habit that lands on zero beats a typical respondent. A named-hit condition was rejected for the headline because naming one of three claims correctly is harder than sensing that something is wrong and would have dragged the typical respondent toward or below zero. Naming is instead its own second reading, reported only on the naming exercises and never averaged into d-prime: on the seven exercises that carry a defect, each answer is classed as named, sensed but misnamed, or missed, and the sensed-but-misnamed count is printed because it separates a person who feels something is off from a person who can say what it is. The two error directions, a defect let through and a comment spent on an ordinary line, are counted and printed separately and never netted. A declared cost matrix prices a let-through at six units and a false alarm at three, in words a let-through costs twice a false alarm; it is a value judgement and is printed on the report above the two counts. It was calibrated against the two fixed habits on this file: flagging everything costs 111 units, flagging nothing 198, and the prior-drawing respondent about 105, so neither habit beats a typical respondent at the printed prices. Every option, pair, position and value carries a declared prior authored before anybody sat the instrument: the share of respondents expected to tick, choose or land on it. On the sweeps it is a selection prior per line and on the naming exercises a prior per claim. Zero on every corrected figure is what a respondent drawing from those priors scores, and observed shares will replace the priors once live data exist. No true-or-false prior in the file is an even split. Each of the five areas is scored on its eight exercises with balanced accuracy on a sweep, keyed accuracy on a naming or true-or-false exercise, position matches on a matching or ordering exercise and banded proximity on an estimate, each corrected against its declared prior, and carries an assumed McDonald's omega from an assumed inter-item correlation of .24 and an assumed score standard deviation of 24, both stated as assumptions. Every threshold decision is taken on the unrounded omega. An area with fewer than eight answered exercises or an omega below .70 carries a three-way placement and no figure, and the refusal is printed where the figure would have been. Every reported figure carries its 68 and 95 per cent bands. An unanswered exercise leaves the numerator, the denominator and the chance term together; it never scores zero. An empty sitting scores exactly zero on every figure. A sitting with fewer than twenty-four of the forty exercises answered is refused a report and the refusal is printed in the headline's place. No percentile appears anywhere, because there is no norm group yet. A careless-responding flag count is computed for the operator and never shown to the respondent as a judgement. Two pairs of naming exercises probe the same rule through different situations and their within-pair consistency is printed. The rules stated in the exercises are drawn from the public structure of a dual goods-and-services tax as enacted in India: place of supply for goods and services, forward and reverse charge, the conditions and blocks on input tax credit, the tax invoice, bill of supply, credit note and debit note, exempt, nil-rated and zero-rated supplies, the value of a supply, composite and mixed supplies, and reconciliation between books, the filed return and the auto-drawn statement of supplier reports. Sources drawn on: the Central Goods and Services Tax Act, 2017 and the Integrated Goods and Services Tax Act, 2017 as enacted by the Parliament of India, for the public structure the exercises describe, cited as structure and never as a key to any current rate, threshold or date; Ebrill, Keen, Bodin and Summers, The Modern VAT (International Monetary Fund, 2001), on invoice-credit mechanics, exemption against zero-rating and the credit chain; Bird and Gendron, The VAT in Developing and Transitional Countries (2007), on dual and sub-national value-added taxes; Keen and Lockwood (2010), The value added tax: its causes and consequences, Journal of Development Economics; Keen and Smith (2006), VAT fraud and evasion: what do we know and what can be done?, National Tax Journal, on invoice mismatches and missing-trader patterns; Pomeranz (2015), No taxation without information: deterrence and self-enforcement in the value added tax, American Economic Review, on the paper trail that reconciliation checks; Rao and Chakraborty (2013), Revenue Neutrality of Goods and Services Tax in India, National Institute of Public Finance and Policy, on the design of India's dual tax; Green and Swets, Signal Detection Theory and Psychophysics (1966) and Macmillan and Creelman, Detection Theory: A User's Guide (2005), for d-prime, the criterion and the log-linear correction; Hautus (1995), Corrections for extreme proportions and their biasing effects on estimated values of d-prime, Behavior Research Methods, for the plus-one-half correction; Swets, Dawes and Monahan (2000), Psychological science can improve diagnostic decisions, for reporting both error directions under unequal costs; Cohen (1960), A coefficient of agreement for nominal scales, and Brennan and Prediger (1981), for chance correction against declared marginals; McDonald, Test Theory: A Unified Treatment (1999), for omega; Gollwitzer and Sheeran (2006), on implementation intentions; and Haladyna, Downing and Rodriguez (2002), for the item-writing rules. All exercises are original works written for this instrument. No real business, supplier, tax officer, software product or filing portal is named anywhere in any candidate-facing text, and no statute, section number, notification, form name or portal is the key to any exercise.