Typing Curriculum Audit Checklist: Review a Lesson Step by Step

By Evan

To audit a typing lesson, compare its promised skill with the input, actions and feedback learners actually encounter. Record the evidence for each finding, choose the smallest suitable fix, and try the revised task again. You can begin with a sample lesson in a browser; access to source code is useful for deeper checks but is not a prerequisite for a classroom review.

This checklist is for teachers and designers reviewing a lesson for a defined learner group. It uses a constructed example throughout, then provides a worksheet you can copy for your own review. The checklist is an author-proposed review method, not a certification scheme.

Audit one lesson with the worksheet.

Define what you are reviewing

Write down the lesson or sample, review date, learner group, keyboard layout, device and promised skill. A useful scope might be: “One introductory C/G/N lesson for adults on a physical US QWERTY keyboard, reviewed for opportunities to practise all three letters.” That is narrower and more defensible than judging a whole course from one screen.

If the goal is functional computer use, include the needed actions as well as character entry. Northstar’s Basic Computer Skills standards include knowledge of special keys and navigation keys. The framework helps identify relevant actions, but it does not specify this checklist’s pass conditions.

A classroom review versus a source audit

In a classroom review, record what the learner sees, enters and receives as feedback. You can discover missing target characters or unclear instructions without reading code. Mark anything you cannot inspect as “not observed.” One sample may not reveal a later review lesson or an optional setting.

With source access, also inspect how stored text becomes actual input, how actions are handled and what the score counts. A source record can explain a behavior, but still try the relevant flow in the browser. Formatting in a file and a required learner action are not necessarily the same thing.

Collect the prompt, instructions and result

Keep the exact displayed instructions and prompt alongside your observation. A title such as “Practise G” is an intention; the prompt supplies the opportunities. For a copy task, count what the learner is required to enter, including meaningful spaces, case, digits and punctuation.

For editing, collect the starting text, cursor or selection, required action and expected result. If a task says to use a particular shortcut, correct final text alone does not prove that shortcut was used. A learner could reach the same result with a mouse or by retyping.

Here is an illustrative editing specification: start with Annx|, where the bar shows the cursor; use Backspace, then type a; expect Anna|. Observe both the action and result. This is a task to review in a text editor, not a claim that every typing player supports it.

Walk through a sample audit

Imagine a lesson titled “Combine C, G and N.” Its stated purpose is to practise all three letters among familiar lowercase keys. One screen asks the learner to copy:

can can can

This is a constructed example, not an observation about a current Typekute lesson. Count the letters: C appears three times, N three times, and G zero times. There are nine non-space characters. C/G/N target letters account for six of them, so target share is 6 ÷ 9 × 100 ≈ 66.7%. The denominator excludes spaces for this particular coverage calculation; it is not an accuracy score.

The combined percentage hides the useful finding: there is no G opportunity. You do not need a universal minimum percentage to conclude that this screen cannot practise all three named letters. Record the finding at screen level, then inspect the other screens before judging the whole lesson.

If the screen’s intended job is controlled combinations, a possible rewrite is c g n g c n. It contains two occurrences of each letter. That fixes this example’s missing-G and balance problem, but six letter entries do not establish sufficient practice or learning. If its job is word practice, use meaningful words containing G instead; the rewrite needs to preserve the task’s purpose.

Finish the audit across a small lesson, not only one line

The following is a completed, constructed review of a three-screen lesson. Its declared purpose is to combine C/G/N, then use them in familiar words. Assume A, O, Space and the required lowercase letters have already been introduced. This assumption is part of the example's scope, not a finding about a live course.

A small lesson reviewed from introduction to application
Screen and intended job	Original prompt and finding	Revised prompt	Specific recheck
1. Combine C/G/N	`can can can`: C=3, G=0, N=3; the named G target is missing	`c g n g c n`: C=2, G=2, N=2	Confirm every target appears, the displayed order matches required input and Spaces are explained.
2. Use targets in words	`can can can`: repeats one word and still supplies no G	`can nag cog`: C=2, G=2, N=2; nine letters, six target letters	Confirm the same 66.7% target share now includes G, and that every other required letter is already introduced.
3. Revisit changed words later	No later screen is included in the example route	`cog can nag`, returned to in another encounter	Record performance and help with the changed order. Recognize that reordering the same words is a limited variation, not a broad transfer test.

The revised second screen has the same combined target share as the original can can can, yet supplies a different opportunity set. This is why a percentage cannot replace per-target counts. The first screen deliberately uses separated letters; the second deliberately uses words. Making both look like ordinary prose would erase their distinct purposes.

Before treating the revision as usable, try it in the actual player. Confirm the prompt is legible, the expected spaces are required, correct input advances, a deliberate wrong letter produces intelligible feedback, and the documented correction action works. These checks establish task behavior. Observing a learner later is a separate check of what they can do after instruction.

Use the learner response to choose the next repair

Suppose the constructed learner response to can nag cog is can naf cog. One incorrect G in a word does not establish a stable weakness or explain its cause. Ask what happened and inspect another appropriate G opportunity. If the learner repeatedly searches for G, return to locating it and then its word transition. If they explain the key correctly but misread the prompt, improve reading support instead.

A different response, cannagcog, points to the separation task. Check whether Spaces were explained and whether the player made missing Spaces visible. Adding more G repetitions would not directly address that finding. If the learner can complete each short word but loses their place in the combined line, try a shorter presentation and then rebuild the line; record that change so the recheck is not mistaken for an identical task.

Keep the intervention small enough to interpret. Changing the text, keyboard, feedback rule and passing threshold simultaneously makes it difficult to identify which change resolved the original problem. You may need several changes for a usable lesson, but record each consequence and recheck the specific promise rather than declaring success from a final score.

A missing Enter is a different problem

A visual line break does not tell you whether Enter is required. In a sample called “Use Enter,” try the task and record whether the learner must create a line break, whether that action is accepted, and whether feedback evaluates it. Pressing Enter to start the next screen is navigation, not evidence of entering a newline in the task.

If Enter input is not required, choose between describing the existing copy task honestly and creating a genuine Enter task. If the environment cannot accept that action, changing the title or adding a newline in the stored text will not establish runtime support.

Check six things that affect the lesson’s purpose

A practical lesson-review checklist
Check	Question	Evidence to record	Possible response
Objective	Does the required task match the stated skill?	Instruction, prompt and actual action	Clarify the objective or rewrite the task
Prerequisites	Are required keys and actions already taught or introduced here?	New requirements and preceding instruction	Introduce the missing component or simplify the prompt
Opportunities and coverage	Does every named target appear in the appropriate task?	Per-target counts, combinations and contexts	Add the missing target or balance the activity
Meaningful task	Does the text serve this activity’s purpose?	Readable words or a clear reason for a character pattern	Replace awkward text or use a more suitable activity type
Later practice and check	Where does the skill return in changed text?	Later task, timing and help allowed	Add a later opportunity; mark uninspected material as not observed
Runtime and feedback	Can the learner perform the promised action, and is it evaluated?	Observed input handling, correction behavior and feedback	Fix instructions or runtime, then repeat the interaction

Use a metric suited to the task. Per-key counts expose a missing letter; word coverage shows where a target appears in words; an editing task needs actions and state. Applying one target-character percentage to every activity would miss these differences.

Also record repeated versus varied opportunities. A prompt of gggg is entirely G, but it offers four repetitions and no transition to another letter. A sentence with a few Gs may be an appropriate application task if its goal is using familiar keys in meaningful text. Neither percentage alone decides instructional quality.

Check what the score actually measures

Read the score definition or inspect what changes after a mistake and correction. Does accuracy count initial incorrect entries, errors left in the final text, or another rule? What enters the denominator? Are corrections allowed, required or blocked? If the definition is unavailable, record that limitation rather than equating similarly named scores.

Soukoreff and MacKenzie (2003) distinguish corrected errors, errors left uncorrected and actions used to fix them. Their framework explains why final text alone cannot reveal all entry errors. A Backspace count is also not automatically a corrected-error rate.

For an illustrative prompt cat, entering cat immediately and entering cax, Backspace, t both finish correctly. The second includes a mistake and repair. If the lesson promises feedback on initial accuracy, check whether that distinction appears in its measurement. If it teaches repair, observing the correction can be relevant evidence.

Dhakal and colleagues’ 2018 typing study separately examined errors and correction behavior in transcription tasks. For your review, keep task conditions and metric definitions attached to each result. A score from copying a visible sentence does not establish composition or editing ability.

Decide what to fix and how to recheck it

Choose a response that fits the actual consequence. A missing target is a content problem; unsupported editing is a runtime problem; an unclear scope can be a wording problem. Treat a failure that prevents the promised task as more urgent than a minor wording improvement, without pretending this priority is a scientific rating scale.

From finding to a specific recheck
Finding	Consequence	Suitable fix	Recheck
No G in the example C/G/N screen	No opportunity to enter G	Rewrite for all targets, or narrow the objective if that is the intended task	Count the revised input and try it
Unknown symbol appears in an introductory capitals task	A new requirement obscures the capitals observation	Introduce the symbol first or split the task	Inspect order and try the revised route
Editing action cannot be performed or evaluated	The promised editing check is unavailable	Change the environment or the promise	Repeat the action and inspect resulting state and feedback
Only familiar text was checked	Performance on changed text remains unknown	Add a later comparable task	Observe learners under the stated conditions

Adding G resolves the missing-G finding. It does not establish retention. A runtime fix needs a new browser interaction; a learning claim needs evidence from learners. State which question your recheck answers.

Use the audit worksheet

Copy the worksheet for one lesson. Include its scope, date, learner group and device above the table. The first row is filled from the constructed C/G/N example; the second is a question-based row for your own review, not an unfinished audit finding.

Lesson audit record
Scope and promise	Observed evidence	Finding	Action	Recheck and conclusion
One example combination screen; practise C/G/N	`can can can`; C=3, G=0, N=3; nine letters	Missing G; this screen cannot practise all named targets	Use `c g n g c n` if the intended job is combinations	Confirm two entries per letter and test the screen; learning remains unmeasured
Which screen and skill are you checking?	What exact input, action and feedback did you observe?	What does that evidence establish, or leave unknown?	Should you retitle, rewrite, split or change the runtime?	What will you try again, and what conclusion will that support?

To choose a real sample, explore the available Typekute lessons. Apply the same questions to the task you encounter, without assuming this article’s constructed example describes it.

What the audit can conclude

A lesson audit can establish that a particular prompt lacks a target, a prerequisite is missing, or a required action was not available in the tested flow. A content or software check can verify a specific repair. Neither alone measures lasting learning or guarantees a speed result.

For a course-wide conclusion, expand and state the inspection scope. For an effectiveness conclusion, define a learner group, outcome and study method. If your next decision is selecting a course, read how to interpret standards and research claims. If it is redesigning the route, use the curriculum planning worksheet.