Secondary 2 English Tuition | When Should You Change English Tutors?

46. Deep Manual | Diagnose the Relationship Before Replacing the Person

Changing tutors is easiest when dissatisfaction has one obvious cause. Most real cases are harder. A learner may be making some progress but not enough. School results may be flat while independence improves. The tutor may be knowledgeable but the group may be too large. The student may dislike the tutor because scaffolds are being removed, or like the tutor because the tutor supplies too much help. A family therefore needs a method that separates evidence from reaction without becoming trapped in endless review.

The governing principle is to diagnose the relationship system. A tutoring relationship contains at least seven moving parts: learner state, active goal, tutor diagnosis, teaching method, practice architecture, delivery/dosage and transfer into school. A change decision should ask which of these is failing and whether the present tutor can repair it. Replace the person only when the person or relationship is materially limiting the repair.

47. Establish the comparison period from evidence gates

Families often ask how many lessons they should “give” a tutor before deciding. A universal session count is misleading because goals vary in complexity. Instead, define gates. Has the tutor seen enough representative work to form a testable diagnosis? Has the learner been taught the proposed mechanism? Has there been enough guided practice to show whether the method is learnable? Has support been reduced enough to test independent production? Has at least one transfer check occurred?

A relationship should not be judged before it has had a fair opportunity to pass the relevant gate. But it also should not remain indefinitely before the same gate. If weeks of lessons cannot produce a clearer diagnosis than “needs practice”, the problem is not merely that learning takes time. The programme has failed to improve its model of the learner.

48. Separate lack of progress from lack of evidence

Sometimes a family believes progress is absent because nobody has looked carefully. The learner’s mark remains similar, but prompt level has fallen, old errors have changed and school feedback has become more specific. In another case the tutor claims strong progress, but all evidence comes from corrected homework and familiar practice.

Before changing tutors, improve the evidence. Compare independent fresh work, old and new school artefacts, prompt levels and one delayed task. If the picture becomes clearer, the decision may change. This step protects both directions: it prevents unfairly discarding an effective tutor and prevents weak tutoring from hiding behind activity.

49. Separate slow learning from static teaching

Some English mechanisms take time to stabilise. Paragraph development, nuanced inference and language control under authentic writing load are not instant repairs. Slow learning can be legitimate when the tutor is generating increasingly precise information, adjusting tasks and showing movement in leading indicators.

Static teaching looks different. The same worksheets, explanations and corrections recur while the learner fails in the same way. The tutor’s description does not become more precise. No new hypothesis is tested. The passage of time is then not evidence of necessary complexity; it is evidence that the instructional loop may be stuck.

50. Separate learner resistance from method mismatch

A learner may resist a method because it exposes a weakness, because it is poorly explained, because the cognitive load is too high or because the method genuinely does not fit. The tutor should not interpret every objection as laziness, and the family should not interpret every objection as evidence that the tutor is wrong.

Test the source. Reduce task size while preserving the same mechanism. Explain the purpose. Offer a different representation. Ask the learner what becomes difficult. If resistance decreases when the method becomes intelligible, the relationship may be fine. If the learner repeatedly cannot access the method and the tutor cannot adapt, fit becomes a legitimate concern.

51. Separate personality discomfort from learning harm

Not every effective teacher feels like a natural personality match. A quieter student may find a direct tutor intense. A highly verbal student may find a reflective tutor slow. Mild discomfort can be part of adjusting to a different intellectual style.

Learning harm is different: chronic humiliation, contempt, fear of attempting, repeated boundary violations, discriminatory treatment or a pattern in which the student hides errors because honesty feels unsafe. The issue is not preference but whether the relationship preserves enough dignity and psychological safety for accurate diagnosis and serious learning.

52. Separate warmth from educational fit

A warm tutor can create trust and persistence. Warmth is valuable. But a relationship may feel excellent while the tutor performs too much of the learner’s thinking, avoids difficult feedback or never tests independent work. Conversely, a tutor may be less socially expressive but highly attentive, respectful and effective.

Evaluate warmth alongside diagnosis, teaching quality, adaptation, transfer and independence. The question is not whether the learner likes the tutor; it is whether the relationship helps the learner become more capable while remaining humane.

53. Separate novelty effects from genuine tutor improvement

A new tutor often creates a short burst of motivation. Materials are different, explanations feel fresh and expectations reset. Performance may improve simply because attention rises. This is not fake; novelty can help. But it is not yet proof that the new system is better.

Compare performance after the novelty period. Does the new tutor diagnose more precisely? Are fresh tasks improving? Is adult labour decreasing? Does school transfer appear? A switch should ultimately be justified by a stronger learning system, not only by the emotional energy of change.

54. Separate relationship knowledge from inertia

Long-term tutors accumulate useful knowledge: which prompts work, which topics expose weakness, what the learner misunderstands under pressure and which routines have already been tried. This is a genuine asset. It should not become a reason to remain forever.

Relationship knowledge is valuable only if it is being used to make teaching more precise. If a tutor has known the student for years but still cannot state the current weak link beyond broad labels, tenure has become inertia rather than expertise.

55. Audit the tutor’s diagnostic resolution

Ask how the tutor describes current performance. Low-resolution descriptions include “weak comprehension”, “bad grammar”, “needs to write more” or “not focused”. Higher-resolution descriptions identify the decision and condition: “selects relevant evidence but overstates causal claims when clues are separated across paragraphs”; “sentence agreement is stable in drills but breaks when complex noun phrases appear in original writing”.

Resolution should increase over time. Early uncertainty is normal. Permanent vagueness after repeated access to work suggests that diagnosis is not accumulating.

56. Audit the tutor’s error model

Does the tutor treat each mistake as an isolated item or group errors into meaningful families? A strong error model distinguishes missing knowledge, wrong strategy, weak retrieval, execution failure, transfer failure and simple noise. This matters because each demands a different intervention.

If every error is answered by correction and repetition, the programme may be too shallow. If the tutor can explain why the same surface mistake can come from different causes, the diagnostic system is more credible.

57. Audit the tutor’s feedback leverage

Count not how many corrections the tutor gives but how many future decisions those corrections improve. High-leverage feedback names a reusable relationship: evidence is related but not direct; the paragraph repeats an earlier function; the sentence changes the certainty of the claim; the revision fixes wording but not logic.

Low-leverage feedback can be accurate yet local. If tutoring produces hundreds of correct corrections with little independent transfer, the problem may be feedback architecture rather than tutor knowledge.

58. Audit the tutor’s use of models

Models should illuminate quality, not replace generation. Ask when models are shown, what learners are expected to extract and how independent production is later tested. If every essay begins by imitating a model introduction, the learner may acquire a style without acquiring the decisions that created it.

A stronger model sequence is attempt, comparison, analysis, changed prompt, independent production. If the tutor can redesign toward this, switching may be unnecessary. If model dependence is treated as the permanent method, concern is reasonable.

59. Audit the tutor’s use of school work

School work should provide authentic evidence and alignment. It should not consume every session. Ask whether school assignments are being used to identify transferable patterns or merely completed and corrected.

A strong tutor can move away from the exact assignment, teach the mechanism on fresh examples and return later to check transfer. A permanently reactive homework service may be the wrong product if the family’s goal is deeper English development.

60. Audit the tutor’s progress measurement

Look for baseline comparison, process evidence, independent fresh tasks, transfer, retention and prompt level. The system does not need a complex dashboard. It needs enough evidence to answer whether the learner has changed and what should happen next.

Be cautious with progress reports built mainly from attendance, chapters covered, worksheets completed or broad comments. Those describe service activity. They do not prove learner change.

61. Audit whether progress changes the next lesson

Measurement should alter instruction. If Iona’s evidence selection becomes stable, the next lesson should not mechanically repeat the same task because it sits next in a workbook. If Maren’s paragraph jobs are strong but qualification remains weak, attention should shift.

A tutor who collects evidence but follows an unchanged sequence may be running a curriculum programme rather than a diagnostic programme. That can still be useful, but families should know which service they are buying.

62. Audit whether failure changes the next lesson

Repeated failure should increase precision. The tutor should ask whether the task was too hard, the prerequisite missing, the explanation weak, the practice insufficient, the transfer jump too large or the goal wrong. More of the same is justified only if the programme can explain why repetition is the missing ingredient.

63. Audit the learner’s cognitive share of the lesson

Who does most of the reading, explaining, deciding, generating, checking and revising? Tutor talk can be necessary during modelling, but over a full cycle the learner should perform increasing amounts of the target cognition.

A useful observation is the ratio between explanation and production. If every lesson contains rich tutor explanation but little learner retrieval or independent application, the service may feel intellectually impressive while producing weak transfer.

64. Audit the learner’s help-seeking quality

A successful tutoring relationship should gradually improve the questions the learner asks. “I don’t understand” may become “I can find the evidence but I cannot decide how strong the claim can be.” “My essay is bad” may become “My examples are relevant but two paragraphs do the same job.”

If help-seeking remains broad after months of tutoring, the student may still depend on the tutor to perform diagnosis. That does not prove the tutor is poor, but it identifies a missing independence goal.

65. Audit the tutor’s capacity to fade

Ask what support has disappeared since the relationship began. Can the student now plan without a template, locate evidence without highlighting, revise without line-by-line correction or complete homework without adult initiation? If the answer is none, investigate why.

Some learners genuinely need sustained support. The warning sign is not continued support itself; it is continued support without a model of what it is doing or what would justify changing it.

66. Audit whether the tutor can say “this goal is complete”

A relationship with no retirement criteria risks permanent expansion. Strong tutors should be able to acknowledge when a weakness is stable enough to stop receiving prime lesson time. They may identify a new need, but it should have its own evidence and rationale.

If success never changes the service, the programme may be structurally biased toward continuation.

67. Decide whether to change goal, method, dosage, format or tutor

Use the smallest change that addresses the diagnosed failure. If the goal is vague, clarify it. If the goal is sound but teaching is not producing learning, change method. If learning occurs but does not consolidate, inspect dosage and retrieval. If the tutor is strong but the environment blocks interaction, change format. If the tutor cannot or will not make the required changes, then change tutor.

This sequencing protects relationship knowledge while preventing loyalty from becoming inertia.

68. Run a controlled method change

When the issue appears methodological, change one important element and define the evidence. If verbal explanation has not produced paragraph development, move to contrastive examples and learner-generated structures. If full worksheets have hidden evidence-search weakness, use shorter targeted passages and think-aloud diagnosis.

A controlled change lets the family see whether the tutor can adapt and whether adaptation matters. Changing several variables at once may feel decisive but makes learning difficult to interpret.

69. Run a controlled dosage change

If the target is correct and teaching works during lessons but retention is weak, the programme may need more frequent retrieval rather than more explanation. This can mean shorter between-lesson practice instead of another full tuition session.

If the student is overloaded or performs well independently between lessons, reduce dosage and test whether capability holds. A tutor who interprets every dosage question as a reason for more paid hours deserves scrutiny.

70. Run a controlled format change

If group interaction is useful but individual practice is insufficient, move temporarily to a smaller group or add periodic individual diagnostics. If one-to-one has become overly tutor-dependent, introduce structured peer comparison or more independent blocks. If online delivery creates distraction, test in-person conditions or a different online architecture.

Format should serve the current learner job. There is no universally best delivery mode.

71. Run a controlled transfer test before switching

If tuition work looks strong but school work does not, design one fresh task that reproduces the suspected school difference: tighter time, unfamiliar wording, mixed question types, reduced scaffolds or changed genre. This reveals whether the problem is transfer rather than teaching acquisition.

If the tutor can identify and repair the transfer boundary, switching may be unnecessary. If the tutor denies the gap or has no method for investigating it, the concern strengthens.

72. Define what would count as a successful repair of the current relationship

A pre-switch adjustment needs its own success criteria. For example: within the next review cycle, the tutor will define one active writing goal, use two fresh independent tasks, reduce line-by-line correction and provide a short evidence review. The family is not demanding a guaranteed grade. It is testing whether the system can become targeted and transparent.

73. Define what would count as a failed repair

A failed repair is not simply that final performance remains imperfect. It is that the agreed structural changes do not occur or generate no better information. The tutor continues broad worksheet practice, heavy homework correction and vague reporting despite the review. At that point, changing tutor is a decision about programme responsiveness rather than impatience with learning speed.

74. Build a portable learner record before switching

Collect a small pack: one baseline reading sample, one recent independent reading sample, one early and one recent writing sample, current school feedback, active error families, current goals, prompt level and a note about methods tried. The record should fit inside a manageable folder rather than becoming an archive of every worksheet.

This portable record prevents the next tutor from spending months rediscovering known information and allows them to challenge previous assumptions with actual evidence.

75. Treat the switch as a hypothesis, not a magic reset

A new tutor should not be expected to erase every difficulty through novelty. State why the switch is being made: stronger diagnosis, better interaction, smaller group, more transfer testing, clearer communication, different expertise or another explicit reason. Then review whether the new system actually provides that improvement.

This protects families from serial switching driven by disappointment. Change should create a better educational architecture, not simply reset expectations.

76. Tutor-Change Clinic | Forty Secondary 2 Decisions

The scenarios below stress-test the change decision. Each asks the same underlying question: what evidence would distinguish a tutor problem from a goal, method, format, dosage, transfer or normal-variation problem? The answer is deliberately not always “switch” or “stay”. Good decisions often begin with one controlled test.

77. One examination score drops sharply

Do not switch from one result. Compare paper difficulty, question mix, timing, topic familiarity and the learner’s error families. If independent tuition checks and earlier school work were improving, investigate the assessment before blaming the relationship. If the paper exposes the same old weakness that tuition claims to have fixed, use it as a transfer test. A strong tutor should be able to explain the discrepancy and adjust the programme.

78. Marks stay flat for a term

Flat marks are ambiguous. Ask whether later work is harder and whether leading indicators changed. Fewer prompts, better paragraph architecture, more precise evidence and stronger completion can precede mark movement. Concern rises if marks are flat and the tutor cannot identify any changing mechanism, transfer evidence or next hypothesis. The absence of a grade increase is not itself the diagnosis.

79. Marks rise but independence falls

This pattern deserves attention. Determine whether the tutor is supplying more hidden labour: planning, wording, checking, evidence selection or homework rescue. If the student’s products improve only because adult support has intensified, ask the tutor to run protected independent checks and fade one support. If the relationship resists any move toward learner control, changing tutor may be appropriate despite higher marks.

80. The learner loves the tutor but progress is vague

Preserve the relationship if it can become more diagnostic. Ask for one active goal, baseline evidence and a fresh-task review. Warmth is an asset, especially if it supports honest attempts. It is not a substitute for targeted instruction. If the tutor can add evidence without damaging the relationship, staying may be best. If the service remains primarily companionship and homework supervision while being sold as remediation, expectations should be reset or the tutor changed.

81. The learner dislikes the tutor but independent work improves

Find out what “dislike” means. The tutor may be appropriately reducing reassurance, or the interaction may be disrespectful. Compare dignity, challenge and learning evidence. Productive discomfort can coexist with progress; chronic fear or humiliation should not be rationalised as rigour. Invite the learner to describe specific moments and decide whether interaction can be repaired without weakening educational standards.

82. The tutor is excellent but the group is too large

Change architecture before changing person if possible. If Leonie needs repeated individual timed production and receives little opportunity because the group is large, a smaller group or periodic one-to-one diagnostic block may solve the problem. The relevant measure is whether the new format increases learner production and targeted feedback. Tutor expertise cannot compensate indefinitely for a structure that prevents the required interaction.

83. One-to-one tutoring becomes too dependent

The intimacy of one-to-one can encourage constant confirmation. Ask the tutor to insert independent blocks, delay reassurance and require learner diagnosis before feedback. If independence improves, the tutor was not the problem; the interaction design was. If the tutor believes constant intervention is the value proposition and refuses to fade, a change may be needed.

84. Online lessons are full of distraction

Test delivery before expertise. Reduce notifications, change device position, shorten direct-instruction segments and increase active production. If attention and work quality improve, keep the tutor. If online conditions remain structurally poor for this learner, move in person or to another online architecture. A format mismatch should not be misdiagnosed as weak English teaching.

85. In-person travel makes the learner exhausted

A strong lesson can lose value if travel cost consumes sleep, homework time or attention. Compare total weekly cost, not lesson quality alone. The right change may be online delivery, lower frequency or a closer provider. If the existing tutor can adapt format, relationship knowledge can be preserved. Educational effectiveness includes the learner’s total capacity.

86. The tutor uses many worksheets but the student likes them

Worksheets are not the problem. Ask whether they are selected from a diagnosis, whether errors change future practice and whether skill transfers beyond familiar formats. If yes, worksheet-heavy teaching can be effective. If the sequence is simply “finish book, mark book, start next book,” ask for a targeted cycle. Change tutor only if the programme cannot become responsive.

87. The tutor rarely uses worksheets

Absence of worksheets is not proof of sophistication. Rich discussion, modelling and authentic texts still need independent production and progress checks. If every lesson feels insightful but the learner cannot reproduce the thinking alone, ask for fresh performance tasks. Judge output and transfer, not material style.

88. Every essay returns heavily rewritten

This is a serious authorship and measurement problem if it is routine. Ask the tutor to preserve first drafts, limit feedback to current goals and require the learner to make revisions. Then compare future independent writing. If the tutoring model depends on adult rewriting as the core service, change may be justified because learner capability cannot be located clearly.

89. The tutor gives very little written correction

Quantity is not quality. The tutor may use high-leverage oral diagnosis and learner self-correction. Ask whether recurring errors are identified, whether students understand them and whether future work improves. Sparse feedback that transfers can be stronger than pages of red ink. Sparse feedback with no evidence trail is different.

90. The tutor changes goals often

Frequent goal changes can show responsiveness or lack of depth. Ask whether each change follows evidence. A new diagnosis after a revealing transfer test is rational. Switching from vocabulary to writing to comprehension because each week brings different homework is drift. The programme should maintain a stable high-leverage job long enough for learning to consolidate unless evidence falsifies it.

91. The tutor never changes goals

Stability can be useful, but goals should retire, refine or change as the learner changes. If “improve comprehension” remains the goal for a year with no more precise description, the target is probably too broad. Ask for current evidence and exit criteria. Failure to update despite clear progress or contradiction is a warning sign.

92. The tutor says the child needs more practice every month

Ask what kind of practice, which mechanism, what changed under current practice and what result would show enough. Repetition is legitimate when retrieval or automaticity is the bottleneck. It is weak when “more practice” replaces diagnosis. If specificity does not improve after a direct review, the programme may be static.

93. The tutor blames school questions for poor transfer

School tasks can genuinely differ. A strong response is to compare formats and train the cue gap. A weak response dismisses external evidence because it contradicts tuition performance. The learner ultimately needs to operate in authentic conditions. If the tutor refuses to investigate those conditions, change becomes more defensible.

94. The tutor blames the learner’s motivation

Motivation matters, but “not motivated” is a low-resolution explanation. Ask what behaviour is occurring: delayed start, avoidance of unfamiliar texts, refusal to revise, failure to practise. Then ask what conditions affect it. A tutor who tests task design, challenge, ownership and success evidence is diagnosing. A tutor who uses motivation as a permanent explanation may be avoiding instructional responsibility.

95. The student is motivated only before tests

This may be a goal-design issue rather than tutor fit. Connect ordinary practice with visible learner goals and use smaller evidence cycles. If the tutor can create relevance and self-monitoring, keep the relationship. If every lesson depends on external pressure and no independence system develops, reconsider the programme architecture.

96. The tutor promises a guaranteed grade increase

Be cautious. Tutors can control teaching design, not every assessment condition. Ask what mechanisms the tutor expects to change and how progress will be measured before marks. Overconfident guarantees can signal weak uncertainty management. Prefer tutors who can make strong instructional commitments without pretending to control outcomes they cannot guarantee.

97. The tutor says improvement cannot be measured

English is complex, but complexity does not make measurement impossible. Independent samples, error patterns, prompt levels, transfer and school work provide useful evidence. If the tutor rejects all measurement, the programme cannot distinguish learning from activity. A change may be justified if this stance is fundamental.

98. The tutor measures everything

Over-measurement can also be weak design. If lessons generate dashboards but little teaching, ask which measures change decisions. A small evidence bundle is usually enough. The issue is not whether data exist but whether the programme uses them intelligently.

99. The learner’s confidence falls

Compare confidence with capability. The learner may be seeing complexity more accurately or may be experiencing damaging interaction. Ask whether independent performance is improving, whether mistakes are treated productively and whether the learner can explain the current challenge. Falling confidence with stronger calibration is different from fear, shame or learned helplessness.

100. The learner’s confidence rises but accuracy does not

Warm support may have improved comfort before skill. That can be useful, but the tutoring goal should now connect confidence to evidence. Ask the learner to predict performance and justify decisions. If the tutor treats confidence as the final outcome while academic capability stays static, expectations may be mismatched.

101. The learner stops asking questions

This can mean increased independence or reduced safety. Look at performance and behaviour. Does the learner diagnose and act without help, or do they hide uncertainty? A strong relationship makes fewer questions more precise, not merely less frequent. If silence comes from fear of correction, fit is poor.

102. The learner asks better questions but grades are unchanged

Improved help-seeking is valuable process evidence. Preserve it while checking whether the underlying mechanisms are moving. If task difficulty has risen, flat grades may coexist with stronger self-regulation. The programme should still eventually connect process gains with authentic performance.

103. The tutor does not communicate often

Frequency alone is not the standard. Some families need only periodic concise reviews. Ask whether communication is sufficient to understand goals, progress and concerns. A low-message relationship with strong clarity can be excellent. Chronic opacity despite requests is different.

104. The tutor sends detailed reports every week

Detailed communication can be useful, but assess content. Are reports describing learner change, transfer and next decisions, or mainly listing completed activities? Administrative sophistication should not be mistaken for instructional sophistication.

105. The tutor and school teacher use different methods

Difference is not automatically conflict. Ask whether both methods express compatible underlying principles and whether the learner can operate successfully in school. A good tutor helps the student translate across valid methods. If the tutor dismisses school requirements categorically, the learner may be caught between systems.

106. The tutor disagrees with school feedback

Disagreement can be legitimate. Compare the rubric, task and actual work. The tutor should explain the evidence, not simply assert superior expertise. When school feedback is ambiguous, the learner still needs to satisfy explicit requirements while understanding broader principles. Professional disagreement should improve clarity, not recruit the student into an adult conflict.

107. The student needs a different kind of expertise

A tutor may be excellent at foundational repair but less suited to advanced stretch, or strong in writing while another specialist is better positioned for a specific oral-performance need. Changing or adding expertise can be rational without implying failure. The current job has changed. Use a handover so previous learning is preserved.

108. The tutor is strong but the learner has outgrown the service

If original goals are complete and the service has no meaningful next job, leaving is success. Do not invent a deficit to preserve continuity. The learner may move to independent work, stretch with another specialist or simply stop formal tuition. A good tutor can acknowledge this without defensiveness.

109. The family changes tutors repeatedly after every disappointing test

Serial switching destroys diagnostic continuity and can teach the learner that difficulty belongs to the provider rather than to a process that needs training. Establish evidence gates and commit to a fair cycle unless there is a professional reason for immediate exit. The family needs a better decision system, not merely a new tutor.

110. The family remains for years because the tutor “knows the child”

Relationship knowledge matters only if it sharpens the programme. Ask what the tutor now knows that changes teaching and what support has faded over time. If the learner’s description remains generic and the service unchanged, familiarity may be substituting for progress.

111. A new tutor criticises everything the old tutor did

Be cautious of total resets. Previous teaching may contain useful routines and real gains. A strong incoming tutor identifies what should be preserved, what needs testing and what evidence supports change. Blanket criticism can waste accumulated learning and may reflect marketing more than diagnosis.

112. A new tutor agrees with everything in the handover

Agreement can be correct, but important claims should still be verified. The new tutor should see fresh work and form an independent model. A handover is a head start, not a script. Strong continuity preserves knowledge without inheriting untested assumptions.

113. The learner’s workload is becoming unsustainable

The problem may be dosage rather than tutor quality. Audit total weekly load, travel, homework and sleep. Reduce frequency or between-lesson work and test whether outcomes hold. If the tutor insists that more hours are always the answer despite overload, fit is poor because the programme ignores the learner’s full system.

114. Tuition is no longer producing new information

If every review repeats the same description and tasks no longer reveal anything new, ask whether the original goal is complete or the programme is stagnant. A successful mature relationship may need lower frequency or an exit. A stagnant one needs redesign or change. The difference is whether the learner can now perform independently.

115. The learner asks to change tutors

Take the request seriously without making it automatically decisive. Ask for specifics: what helps, what blocks thinking, what feels unfair, what feels too easy, what support creates dependence. Compare the account with work and observed interaction. Secondary 2 learners can contribute meaningful fit evidence. The strongest decision combines learner voice with educational evidence, not one instead of the other.

116. The Pre-Switch Operating System

A high-quality change decision can be run as a short operating sequence. First define the current English job. Second collect representative evidence. Third identify whether the failure sits in diagnosis, method, dosage, format, transfer, interaction or professional reliability. Fourth decide whether one controlled repair is appropriate. Fifth define what that repair must produce. Sixth either continue with a redesigned relationship or switch with a portable handover.

The sequence is deliberately finite. It protects tutors from being discarded after normal difficulty, and protects learners from remaining indefinitely in a system that does not respond to evidence.

117. Build a one-page pre-switch audit

1. What is the current goal?
2. What baseline showed this goal mattered?
3. What has changed?
4. What has not changed?
5. What support does the tutor still provide?
6. Where has transfer succeeded?
7. Where has transfer failed?
8. What adjustment has been attempted?
9. What evidence would justify staying?
10. What evidence would justify switching?

If several answers are impossible because the programme has no baseline, no defined goal and no transfer checks, that absence is itself evidence. The audit is not a bureaucratic requirement; it is a test of whether the relationship can explain its own educational logic.

118. Build a one-page learner voice audit

1. What can you do now that you could not do before?
2. What does the tutor still do for you?
3. Which explanation or routine actually helps?
4. When do you become confused?
5. Can you disagree or ask questions safely?
6. Which work feels too easy or too hard?
7. What happens when the tutor is not present?
8. If you changed tutors, what would you want the new person to keep?

The learner’s answers are evidence, not a referendum. A student may prefer easy work or dislike necessary challenge. The value of the audit is that it reveals the lived mechanics of the relationship, including dependencies adults may not notice.

119. Build a one-page parent audit

Parents should ask what they observe outside lessons. Is homework requiring less intervention? Are school comments changing? Does the learner explain their own weaknesses more precisely? Is total workload sustainable? Are there new conflicts or anxieties? Does the tutoring arrangement consume increasing family coordination?

Parent evidence is strongest when it describes behaviour rather than assigning motives. “Needs two reminders to start instead of six” is more useful than “more responsible”. “Can explain why the answer is uncertain” is more useful than “more confident”.

120. Compare old and new tutors with a shared baseline

If a switch occurs, preserve one or two comparable tasks so the family can evaluate the new system without relying on novelty. The new tutor should not be judged by whether the first lesson feels better. Compare diagnostic precision, learner production, support level and transfer over a meaningful cycle.

The purpose is not to run a competition. It is to learn whether the reason for switching was actually addressed.

121. Avoid a total curriculum reset

New tutors often have preferred materials and sequences. That is reasonable. But restarting from page one can waste prior learning. The incoming tutor should identify what is already stable, what needs verification and what remains genuinely weak.

Maren should not relearn basic paragraph structure if the evidence shows that architecture is secure and the active problem is qualification. Iona should not repeat literal comprehension drills if her issue is cross-paragraph inference. Leonie should not receive generic time-management worksheets if the bottleneck is decision closure.

122. Preserve successful routines across the switch

A change in tutor does not require a change in every routine. If the learner has a useful evidence-scope check, planning method, error ledger or closure rule, preserve it unless the new tutor has evidence that it causes a problem. Continuity lowers switching cost and respects the learner’s existing competence.

123. Retire routines that became tutor-specific

Some routines exist only because the old tutor created a particular worksheet language or sequence of prompts. The new tutor should test whether the underlying skill can survive without those cues. If not, teach the learner to recognise the real task demand rather than memorise another tutor-specific trigger.

124. Use the first new-tutor month as a verification period

The incoming tutor needs enough time to verify the old diagnosis, observe independent work and establish interaction patterns. During this period, resist pressure to prove value through excessive novelty or immediate grade claims. The strongest early result may be a more accurate model of the learner.

At the end of the verification period, the new tutor should be able to say what they agree with, what they have revised and what evidence supports the next goal.

125. Look for better questions, not merely better answers

A stronger tutor often changes the learner’s questions. Maren begins asking whether two paragraph ideas overlap. Iona asks how certain the passage permits her to be. Leonie asks whether another check is worth its time cost. These questions indicate that the learner is internalising criteria.

If the new relationship simply supplies answers faster, the switch may have improved service convenience without improving independence.

126. Look for a shrinking support footprint

Track what the tutor must do for the student. Over time, direct modelling should become cues, cues should become self-prompts, and some self-prompts should become automatic routines. Not every support must disappear, but the overall footprint should reflect the learner’s growing capability.

A tutor change is successful when the new system creates more learner control, not simply a different style of adult control.

127. Look for higher-quality errors

A learner can improve even while still making mistakes. Early errors may show complete misunderstanding; later errors may be narrower, more sophisticated and easier to repair. The incoming tutor should be able to describe this movement rather than using error count alone.

If the old tutor corrected basic errors and the new tutor exposes higher-level reasoning failures, the learner may initially appear worse. Compare the quality and difficulty of the errors before judging the switch.

128. Look for transfer beyond the new tutor’s materials

A new method can produce quick gains on the tutor’s own tasks. Test school work, changed genres and delayed retrieval. The new tutor should welcome this because transfer validates the method. If performance remains locked inside the new materials, the family may simply have traded one tutor-specific environment for another.

129. Know when a second switch is justified

A second switch should not be taboo. If the new tutor fails for the same structural reasons—no diagnosis, heavy adult authorship, no transfer, poor fit—the family should act. But examine whether the family’s own decision process is contributing: unrealistic promises, constant demand for polished homework, changing goals after every test or refusing enough time for transfer.

Improving the support system may require changes on both sides of the tutoring relationship.

130. Know when not to replace tutoring at all

Sometimes the correct conclusion is not “find another tutor” but “the learner no longer needs this kind of tutoring”. If goals are stable and independent performance is strong, replacing one provider with another can preserve dependence unnecessarily. The next page in this lane addresses that exit decision directly.

131. The switch decision for Maren

Before switching, Maren’s family should ask whether the current tutor can move from rewriting essays to teaching independent architecture. If the tutor agrees, preserves first attempts, uses fresh prompts and reduces generated wording, run the redesign. If the tutor’s method fundamentally depends on producing polished final essays together, a tutor whose model prioritises independent writing may be the better fit.

132. The switch decision for Iona

Iona needs a tutor who can distinguish evidence, inference and scope across varied texts. If the current tutor recognises that school transfer is weak and redesigns around changed genres and reduced cues, keep the relationship long enough to test. If the response is simply more familiar comprehension worksheets, the central need remains unmet.

133. The switch decision for Leonie

Leonie’s family should separate execution design from personality. A tutor who insists she is “too slow” without measuring where time leaks may not be diagnosing. A tutor who can show that repeated reopening of closed answers causes most loss and trains a closure rule is addressing the mechanism. If the environment itself makes Leonie afraid to commit, relational fit also matters.

134. The portable handover standard

A world-class handover should allow the next tutor to begin with the learner’s accumulated evidence but not its inherited assumptions. Include current goals, retired goals, representative samples, prompt levels, transfer successes, active error families, methods tried and one unresolved diagnostic question. Leave out unnecessary personal commentary and labels.

The handover is successful when the next tutor can ask better questions sooner.

135. The final pre-switch test

Before an ordinary pedagogical switch, ask four questions. Can the current tutor describe the real problem? Can the tutor change the system in response to evidence? Does the learner have a fair opportunity to demonstrate transfer and independence? Would another tutor plausibly improve the part that is failing?

If the first three answers are yes, switching may add cost without solving anything. If the answers are repeatedly no—and the relationship has had a fair opportunity to repair them—the case for change becomes strong. The decision is then not a judgement of the tutor’s worth as a person. It is a judgement about whether this support system still fits this learner’s next English job.