Reviewer3 (AI review)

November 5, 2026

AI Review (reviewer3.com)

Overview

Role of Academic Adjustment in Student Retention

Contribution

This study demonstrates that academic adjustment serves as the primary statistical link between administrative support for guidance and counselling and students' intentions to remain enrolled.

Gap Between Administrative Encouragement and Resource Allocation

Contribution

The research reveals an institutional discrepancy in Ghanaian schools where leadership encouragement for counselling is strong, but essential practical resources—such as private counseling spaces, protected timetable slots, and workload relief—remain inadequate.

Comments (20)

1. Contradictory institutional level and incompatible age brackets — Critical

transparent sensitivity comparison. The guiding question for this literature is, “how are perceived guidance - and - counselling administrative support, academic adjustment and intended retention related among student respondents in the available Ghanaian pre - tertiary schools?” Two other

The manuscript frames the study around 'Ghanaian pre-tertiary schools' (secondary/high schools), yet the sample comprises 'lecturers or tutors (n = 58)'—terminology typically associated with tertiary institutions—and the demographic questionnaire (Section A, Question 2) begins at age '18 - 24 years' with no option for respondents under 18, whereas the majority of Ghanaian pre-tertiary students are under 18. The authors must clarify the exact type and number of institutions sampled and explain how pre-tertiary students were captured with an adult-only age scale.

2. Unreported item-level statistics and challenge frequencies — Minor

moderate levels of academic adjustment and intention to remain in school.A closer examination of the administrative support items revealed differences in the support provided by schools and that s tudents expressed greater satisfaction with school leadership's encouragement of guidance and counselling programmes than with the availability of private counselling offices and dedicated counselling time. This suggests that although school leaders may recognise the importance of guidance and counselling, the practical arrangements required for effective service delivery remain inadequate. The reported challenges further support

The Findings and Discussion sections make specific quantitative claims comparing individual GCAS items and identifying the most frequently selected barriers from Section E (Question 20) of the questionnaire. However, no item-level means, standard deviations, or frequency counts/percentages for the Section E checkboxes are reported anywhere in the text or tables to substantiate these findings.

3. Absent sample size determination and power analysis — Major

This study used a quantitative cross - sectional mode for analysizing the responses gathered using google forms .Participants were recorded as students (n = 65; 40.4%), lecturers or tutors (n = 58; 36.0%), and supporting staff (n = 38; 23.6%). Because adjustment and retention items refer explicitly

The Methodology section provides no justification for the sample size (N = 161 total; n = 65 students), no description of the sampling frame or recruitment procedure, and no power analysis specifying the target effect size, alpha level, and statistical power for the primary student-only analyses.

4. Duplicate items in five-item academic adjustment and retention scales — Major

With a calculated arithmetic mean for each five - item block , Guidance and Counselling Administrative Support (GCAS) (items on private office space, time, materials, leadership support and coordinator training); academic adjustment (workload, study time, difficult subjects and examination pressure); and intended retention (completion intention, belonging, thoughts about leaving and attendance motivation). Higher scores indicate more favourable perceptions.

Although the Academic Adjustment and Intended Retention composites are described as five-item scales, the text lists only four distinct item domains for each construct, and the appended questionnaire reveals that Item 13 is a verbatim duplicate of Item 12 ('I feel confident handling difficult subjects or topics after receiving guidance from my teachers or the coordinator. 2') and Item 17 is a verbatim duplicate of Item 16 ('I feel a strong sense of belonging to this school that motivates me to continue my studies here. 2'). Averaging five items where one item is duplicated double-weights that specific facet in the composite score and artificially inflates Cronbach's alpha.

5. Omitted primary student-only regression table and correlation matrix — Editorial

This paper examined sample composition, composite means and standard deviations, internal consistency using Cronbach’s alpha, Pearson correlations and linear regression. To reduce sensitivity to unequal error variance in coefficient inference, ordinary least - squares models used heteroskedasticity - consistent With Heteroskedasticity - Consistent Covariance Matrix Estimator, Type 3. (HC3) standard errors and 95% confidence intervals. Separate models related

The Methods section announces Pearson correlations and HC3-robust regression models with 95% confidence intervals for the primary student-only sample (n = 65), and the Appendices jump directly from Table 1 to Table 3 ('Mixed-role sensitivity models'). Table 2 (the primary student-only regression results) and the full correlation matrix are entirely missing from the manuscript, leaving the primary regression models without reported standard errors, 95% confidence intervals, or R-squared values for the separate models, and leaving two of the three pairwise correlations unreported.

6. Sub-threshold reliability of primary predictor in OLS mediation — Critical

The reliability analysis indicated that the measures of academic adjustment and intended retention demonstrated acceptable internal consistency. However, the GCAS measure showed relatively weak reliability (Cronbach's α = .55), suggesting that respondents may have perceived different aspects of administrative support in varying ways. Consequently, the findings involving the overall administrative support measure

The primary independent variable (GCAS) has an internal consistency of Cronbach's alpha = .55, which falls well below the standard threshold (alpha >= .70) for acceptable scale reliability, and the Mixed-Role sensitivity analysis relies on scales with alphas of .55, .69, and .67 (Table 1). In ordinary least-squares regression and product-of-coefficients indirect effect models, substantial measurement error in the predictor and mediator attenuates main effects and distorts partial regression coefficients in unpredictably biased ways, invalidating unadjusted OLS inference.

7. Causal attribution to support built into outcome items — Critical

11. I am able to manage my study time effectively because of support I have received in school. 12. I feel confident handling difficult subjects or topics after receiving guidance from my teachers or the coordinator.

Three of the five items in the Academic Adjustment scale (Items 11, 12, and 13) and one item in the Intended Retention scale (Item 19: 'I am motivated to continue attending school regularly because of the support available to me') are double-barreled statements that require respondents to endorse both the outcome and the causal attribution that school support or guidance produced it. Because the predictor construct (GCAS) also measures perceived school support for guidance, this item wording tautologically builds the hypothesized predictor-outcome link into the outcome measures and confounds the regression and indirect-effect estimates.

8. Incomplete ethics, consent, and conflict-of-interest statements — Editorial

The dataset contains anonymous survey fields for the purposes of the present analysis; identifiable names were neither needed nor reproduced in the manuscript.

The manuscript lacks formal institutional ethical approval details, informed consent documentation, and completed declarations, instead retaining editorial placeholder text on pages 8 and 9 ('Funding and ethics/consent statements are to be completed from the original research records'; 'The author must confirm the accurate conflict-of-interest declaration before submission'). Complete ethics approval and consent verification must be provided prior to publication.

9. Indirect effect in Table 3 does not equal product of reported slopes — Editorial

For an exploratory indirect association, the Guidance and Counselling Administrative Support (GCAS) to - adjustment slope was multiplied by the adjustment - to - retention slope from the joint

The 'Statistical analysis' section (page 5) defines the exploratory indirect association as the GCAS-to-adjustment slope multiplied by the adjustment-to-retention slope from the joint model. For the student-only sample in the Abstract, this product holds (0.673 × 0.751 = 0.505, matching the midpoint of the [0.342, 0.667] interval). However, in Table 3 (page 11) for the mixed-role model, multiplying the reported GCAS-to-adjustment slope (B = 0.621) by the joint adjustment-to-retention slope (B = 0.606) yields 0.376 (even accounting for rounding bounds, [0.6205 × 0.6055, 0.6215 × 0.6065] = [0.3757, 0.3769]), which conflicts with the reported point estimate of 0.373.

10. Single-school case study in Abstract contradicts multiple schools in text — Major

arrangements with students’ academic adjustment and intention to remain enrolled is limited.To examine how perceived administrative support for guidance and counselling relates to academic adjustment and intended retention, distinguishing student responses from staff reports this study explores their practical implications through a cross - sectional , with Ghanaian school as case study.The

The Abstract (page 1) describes the research design as using a 'Ghanaian school as case study' (singular), and the questionnaire items (page 13) refer to 'this school'. In contrast, the Introduction (page 2), Findings (page 5, 'differences in the support provided by schools'), and Conclusion (page 8, 'in Ghanaian schools') state that the study examined multiple schools.

11. Product of coefficients in Table 3 does not match reported slopes — Minor

Exploratory indirect association Guidance And Counselling Administrative Support ( GCAS ) × Adjustment 0.373 [0.279, 0.471]

In Table 3, the exploratory indirect association is defined as the product of the GCAS-to-adjustment slope (0.621) and the joint adjustment-to-retention slope (0.606), which equals 0.376, but the table reports 0.373 without explaining whether this discrepancy is due to rounding of unrounded inputs or reporting the bootstrap mean/median instead of the sample point estimate.

12. Primary student regression table (Table 2) is omitted — Critical

Table 1. Composition, composite scores and reliability ... Table 3. Mixed-role sensitivity models (N = 161)

The Appendix jumps directly from Table 1 to Table 3 (the N = 161 mixed-role sensitivity models), completely omitting Table 2 for the primary student-only sample (n = 65). Consequently, confidence intervals, standard errors, and R-squared values for the primary student regressions are missing from the tables, preventing full verification of the main statistical claims.

13. Missing correlation matrix and bivariate statistics for student models — Major

should be interpreted with some caution.The correlation analysis revealed positive and statistically significant relationships among administrative support, academic adjustment and intended retention. Students who

Although the text states that Pearson correlations were positive and statistically significant among all three study variables, no correlation matrix is provided, and the exact correlation coefficients (r) and p-values for GCAS with Academic Adjustment and GCAS with Intended Retention are never reported.

14. Severe measurement error in GCAS biases joint regression coefficients — Critical

intended retention (B = 0.483, p = .002). With adjustment included, the adjustment – retention association remained positive (B = 0.751, p < .001), while the direct support coefficient was not significant (B = −0.022, p = .888). The student

While the text notes that the low reliability of GCAS (alpha = 0.55) warrants caution, in a joint OLS regression predicting retention from both GCAS and Academic Adjustment (alpha = 0.73), differential measurement error severely attenuates the less reliable predictor (GCAS) while shifting shared variance onto the more reliable predictor. The disappearance of the direct GCAS association in the joint model could therefore be an artifact of differential reliability rather than a substantive indirect pathway.

15. Students cannot validly evaluate coordinator training support — Critical

9. GCAS5. The guidance and counselling coordinator in my school receives support to attend training or workshops.

Item 9 of the GCAS scale asks whether the school's guidance coordinator receives administrative support to attend training or workshops, an internal staffing detail that student respondents would not ordinarily be privy to. Including this item in the student composite score introduces guessing and measurement noise, which helps explain the poor internal consistency (Cronbach's alpha = 0.55) and weakens the validity of the primary predictor.

16. Uncontrolled demographic and school-level confounders — Critical

transparent sensitivity comparison. The guiding question for this literature is, “how are perceived guidance - and - counselling administrative support, academic adjustment and intended retention related among student respondents in the available Ghanaian pre - tertiary schools?” Two other

Although demographic data (sex, age, employment status) were collected in Section A, none of these variables—nor school-level clustering—were controlled for in the regression models. Furthermore, whereas the guiding question frames the study around 'pre-tertiary schools', all respondents were aged 18 or older (up to 45+) and staff were classified as 'Lecturers / Tutors', suggesting an unacknowledged mix of tertiary and pre-tertiary contexts that could confound both adjustment and retention intentions.

17. Point estimate of the student-only indirect association is omitted — Major

The student-only bootstrapped indirect estimate was positive [95% percentile interval (0.342, 0.667)].

For the primary student-only bootstrap analysis, the manuscript reports the 95% percentile interval (0.342, 0.667) but omits the point estimate of the indirect effect (which, as the product of B = 0.673 and B = 0.751, should be approximately 0.505) and its bootstrap standard error.

18. No item-level statistics or checkbox frequencies are reported — Major

moderate levels of academic adjustment and intention to remain in school.A closer examination of the administrative support items revealed differences in the support provided by schools and that s tudents expressed greater satisfaction with school leadership's encouragement of guidance and counselling programmes than with the availability of private counselling offices and dedicated counselling time. This

The manuscript claims that respondents rated leadership encouragement higher than private office space and dedicated timetable time, and that shortages of trained coordinators, inadequate facilities, and heavy workloads were the most frequently selected barriers in Section E (Question 20). However, no item-level means, standard deviations, or checkbox frequency counts/percentages are provided anywhere in the text or tables to substantiate these comparisons.

19. Built-in conceptual overlap across scale items confounds the regressions — Critical

11. I am able to manage my study time effectively because of support I have received in school.

Items in both the Academic Adjustment scale (e.g., Item 11: 'because of support I have received in school'; Item 12/13: 'after receiving guidance from my teachers or the coordinator') and the Student Retention scale (Item 19: 'because of the support available to me') explicitly ask respondents to condition their answers on perceived school support. This item wording builds the predictor directly into the mediator and outcome measures, creating tautological overlap that inflates the observed associations.

20. Duplicate survey items double-weight specific sub-constructs and inflate reliability — Critical

12. I feel confident handling difficult subjects or topics after receiving guidance from my teachers or the coordinator. 13. I feel confident handling difficult subjects or topics after receiving guidance from my teachers or the coordinator. 2

In the Appendix questionnaire, Academic Adjustment items 12 and 13 are identical save for a trailing '2', and Student Retention items 16 and 17 are likewise verbatim duplicates. Because the five-item composite scores were calculated as unweighted arithmetic means without adjustment, confidence in difficult subjects and sense of belonging are inadvertently double-weighted (40% of each scale), which also artificially inflates the reported Cronbach's alphas (0.73 and 0.71) and distorts the regression estimates.

xPeerd (AI review)

November 5, 2026

AI Review (xPeerd — DAReview)

Summary

This manuscript examines whether perceived administrative support for school guidance and counselling is associated with academic adjustment and intended student retention in Ghanaian schools, using a cross-sectional survey with 161 respondents, of whom 65 students form the primary analytic sample. The paper’s main strength is that it explicitly restricts its principal inference to student responses because “adjustment and retention items refer explicitly to the respondent’s personal student experience,” while using the mixed-role sample only as sensitivity evidence . The statistical findings are internally consistent with the paper’s main interpretive claim: administrative support is positively associated with academic adjustment and separately with intended retention, but in the joint model academic adjustment remains significant while administrative support becomes non-significant, suggesting adjustment is the more proximal correlate of intended retention . The manuscript is also appropriately cautious in several places about causal overreach and acknowledges that it does not measure actual retention or temporal ordering .

Its main weaknesses are methodological and presentational. The administrative-support scale has weak reliability (Cronbach’s α = .55), which materially limits confidence in the main predictor . The manuscript also contains substantial grammatical, typographic, and phrasing problems throughout, some of which obscure meaning. In addition, ethics, funding, and data-availability statements remain incomplete, which is a serious submission-readiness concern . I do not see strong evidence of fabrication, but I do see signs of rushed assembly and uneven editorial control.

Potential Major Revisions

1. Clarify and tighten the manuscript’s central research contribution

The paper’s title and abstract suggest a focused institutional contribution, but the manuscript sometimes overstates novelty without sufficiently distinguishing its specific contribution from related work. For example, the abstract states that the study examines “how perceived administrative support for guidance and counselling relates to academic adjustment and intended retention” and “distinguish[es] student responses from staff reports” . That is a plausible contribution, but the introduction should frame this more sharply as a measurement-and-inference refinement rather than an entirely new conceptual model.

At present, some passages are too diffuse. For instance, the manuscript says: “the present studies ask a more specific administrative question” and later “The literature contributes an evidence-bounded account” . These formulations blur whether the paper is offering a literature synthesis, a new analytic strategy, or a substantive educational finding. The introduction should explicitly state:

• the exact gap,
• why prior Ghanaian and non-Ghanaian studies do not answer it,
• why student-only analysis changes the interpretation,
• and what is genuinely new here beyond applying standard regression to survey composites.

2. Address the weak reliability of the key predictor more rigorously

A major methodological concern is the low internal consistency of the Guidance and Counselling Administrative Support scale. The manuscript acknowledges: “the GCAS measure showed relatively weak reliability (Cronbach's α = .55), suggesting that respondents may have perceived different aspects of administrative support in varying ways” . Appendix Table 1 confirms α = 0.55 both in the full sample and student-only sample .

This is not a minor caveat; it weakens the interpretability of the study’s central independent variable. A five-item composite combining office space, time allocation, materials, leadership support, and coordinator training may not be unidimensional. The manuscript should therefore do at least one of the following:

• justify theoretically why these heterogeneous elements belong in one scale,
• report item-total correlations or alpha-if-item-deleted diagnostics,
• test dimensionality,
• or reframe GCAS as an index rather than a latent scale.

Without that clarification, the main finding that GCAS predicts adjustment may partly reflect aggregation across conceptually distinct administrative conditions.

3. Reframe the indirect-effect language to avoid quasi-mediation implications

The manuscript is commendably cautious in some passages, but still risks encouraging over-interpretation by readers. It states: “For an exploratory indirect association, the Guidance and Counselling Administrative Support (GCAS) to-adjustment slope was multiplied by the adjustment-to-retention slope from the joint model” and later reports a positive bootstrapped indirect estimate in the abstract . It also writes that the analysis “investigates patterns consistent with a potential pathway rather than testing a confirmed sequence” .

That caution is useful, but the manuscript still presents a product-of-coefficients analysis in a way many readers will read as mediation. The authors should remove any phrasing that implies process, mechanism, or transmission unless they redesign the study longitudinally or experimentally. The sentence “the resulting indirect estimate summarises how the measured associations align within the fitted linear model. It is not an estimate of the causal effect of changing administrative practice” is strong and should be moved earlier and emphasized more prominently .

4. Resolve the mixed-role design problem more transparently

The study combines students, lecturers/tutors, and support staff in the overall survey, while acknowledging that key outcome items are student-worded. This is a potentially serious design incoherence. The methods admit that “adjustment and retention items refer explicitly to the respondent’s personal student experience” and therefore student responses were the “primary analytical population” . Appendix Table 3 further notes that mixed-role estimates “incorporate 96 staff responses to questions framed as students’ personal experience” and “do not represent student-level effects in N = 161” .

This is an important boundary, but it should be treated more critically. If 96 non-student respondents answered student-experience items, the manuscript should explain:

• how they were instructed to interpret those items,
• whether they answered as proxies, observers, or from personal prior experience,
• why such responses are informative at all,
• and whether including them introduces systematic construct contamination.

As currently written, the sensitivity analysis is reported as if it strengthens robustness, but it may simply reproduce broad directional trends using a conceptually mismatched sample.

5. Strengthen reporting on sampling, setting, and generalizability

The manuscript repeatedly refers to “Ghanaian schools” and “available Ghanaian pre-tertiary schools,” but the context of sampling is underspecified in the excerpts available. The abstract says “with Ghanaian school as case study” , which implies perhaps one school, while later sections use broader language. This ambiguity matters. If the data come from one school or a highly localized convenience sample, claims should be narrowed accordingly.

The paper should specify:

• number and type of schools,
• recruitment method,
• response rate if known,
• school level and region,
• inclusion/exclusion criteria,
• and whether the sample is institutional, municipal, regional, or national in relevance.

Without this, the external validity of the findings is impossible to judge.

6. Complete ethics, consent, funding, conflict, and data-availability reporting before submission

The current manuscript is not submission-ready on research-governance grounds. The end matter states: “Funding and ethics/consent statements are to be completed from the original research records” and “A data-availability statement should be finalised after consent and privacy requirements are checked” . It also says: “The author must confirm the accurate conflict-of-interest declaration before submission” .

These are major omissions. A paper involving human survey responses, especially from students, should not be submitted with placeholders for ethics and consent. The manuscript must identify:

• ethical approval status and committee,
• informed consent procedures,
• any permissions from schools or administrators,
• participant protections,
• conflict-of-interest statement,
• funding source,
• and a finalized data-availability statement.

7. Scrutinize measurement quality and item duplication in the questionnaire appendix

The appendix shows apparent duplicated items:

• “I feel confident handling difficult subjects or topics after receiving guidance from my teachers or the coordinator” appears as items 12 and 13, with item 13 marked “2” .
• “I feel a strong sense of belonging to this school that motivates me to continue my studies here” appears as items 16 and 17, with item 17 marked “2” .

If these are accidental duplicates, that raises concerns about instrument construction and scoring. If they are intended parallel items, the wording is not parallel; it is repeated. This may also partly explain reliability limitations or artificially influence composite means. The manuscript should explicitly clarify whether the duplicates are typographic errors in the appendix, duplicated survey items in data collection, or repeated items intentionally used for consistency checks.

Potential Minor Revisions

1. Extensive language and grammar corrections

The manuscript contains many sentence-level problems that require thorough editing. Examples include:

• “This study explores their practical implications through a cross-sectional , with Ghanaian school as case study” in the abstract, which is grammatically incomplete .
• “These conditions are cannot fully be justified as desired outcomes” on page 1, which is syntactically incorrect .
• “Some studies from Ghana also suggests” should be “suggest” .
• “They only indicatied” contains a spelling error on page 1 .
• “This paper on drew analysis on questionnaires responses” is malformed .
• “Most literature works merely measure outcomes of acdemic adjustment” contains a spelling error and awkward phrasing .
• “the need to asks individuals” is grammatically incorrect .
• “quantitative cross-sectional mode for analysizing” should be revised for both word choice and spelling .
• “gathered using google forms” should capitalize Google Forms .
• “This contributes a complementary institutional perspective...” is acceptable, but several surrounding sentences are run-on and require punctuation repair .
• “this underpin the value” should be “underpins” .
• “It also suggest” in the conclusion should be “suggests” .
• “when examined separatelyIt” is missing punctuation/spacing .

A full professional language edit is strongly recommended before submission.

2. Improve stylistic precision and remove awkward repetition

Several passages over-repeat constructs such as “Guidance and Counselling Administrative Support (GCAS)” in ways that reduce readability, especially in the methods and results . The authors should vary sentence structure and avoid redundant restatement of the full term once the abbreviation is established.

3. Standardize terminology and capitalization

The manuscript inconsistently uses:

• “Guidance And Counselling Administrative Support”
• “Guidance and Counselling Administrative Support”
• “Guidance and counselling”
• “GCAS”

Choose one capitalization convention and apply it consistently across text, tables, and appendices.

4. Improve table formatting and appendix presentation

The appendix tables appear difficult to read in their current form, and labels are inconsistent. For example, “Guidance And Counselling Administrative Support (GCAS) , Mean (Sd)” contains awkward spacing and inconsistent capitalization . Table headings should be standardized, decimal alignment improved, and note formatting cleaned.

5. Ensure the abstract accurately reflects limitations

The abstract is relatively balanced, but it should more directly mention the weak GCAS reliability, because that is central to the interpretation of the main predictor .

6. Tighten the conclusion

The conclusion largely repeats earlier discussion points. It should be shortened and focused on:

• what was found,
• what cannot be inferred,
• and what future research should test prospectively .

AI content analysis for this post-2021 work

Estimated percentage of AI-generated content: 28–42%

This is only a stylistic estimate, not proof. I do not see decisive evidence of fabricated references from the manuscript excerpts reviewed; the citations appear plausible and internally consistent . However, I do see several traits often associated with partial AI-assisted drafting or heavy machine-influenced editing:

• repetitive explanatory phrasing,
• syntactically awkward but semantically fluent sentences,
• inconsistent local grammar despite globally coherent structure,
• formulaic transitions across sections,
• and placeholder-like governance statements at the end.

Highlighted potentially AI-influenced sections include:

• the abstract, where sentences are fluent in intent but structurally broken, e.g., “this study explores their practical implications through a cross-sectional , with Ghanaian school as case study” ;
• the literature synthesis on pages 2–3, where generalizations become repetitive and locally ungrammatical, e.g., “Most literature works merely measure outcomes of acdemic adjustment...” ;
• the discussion and implications sections, where recommendations are polished but somewhat generic and templated in tone .

Assessed epistemic impact: moderate. If AI assistance was used only for drafting prose, the main epistemic issue is clarity, not validity. If it also influenced study design description, instrument reporting, or interpretation, then the risk is more substantial, especially where duplicated items, incomplete ethical reporting, and over-smoothed discussion language could obscure underlying methodological weaknesses.

Recommendations

1. Rebuild the manuscript around a sharper and narrower contribution: student-perceived counselling administration, student-only outcome inference, and cautious interpretation of association rather than mechanism.

2. Reassess the GCAS construct statistically and conceptually. If it is multidimensional, either split it into subdomains or present it as an index rather than a reliability-dependent scale.

3. Remove or heavily qualify any wording that may be interpreted as mediation or pathway testing. Emphasize associational interpretation consistently from abstract to conclusion.

4. Provide full reporting on sampling frame, school context, recruitment, and generalizability boundaries.

5. Clarify the questionnaire appendix, especially the duplicated items, and explain whether those duplicates existed in the administered instrument.

6. Finalize ethics approval, participant consent, conflict-of-interest, funding, and data-availability statements before submission. In its current form, the manuscript has a serious compliance gap .

7. Undertake a full substantive language edit by a proficient academic editor. The current level of grammatical instability would likely prejudice reviewer confidence even where the analytic reasoning is acceptable.

8. For future work, a prospective longitudinal design with observed retention outcomes, school-level clustering, and validated counselling-access measures would substantially strengthen the evidential value of the research question raised here.

XSci (AI review)

November 5, 2026

Summary

The manuscript asks whether students who perceive better administrative conditions for school guidance and counselling also report better academic adjustment and stronger intentions to remain enrolled. Using a Google Forms survey of 161 respondents from Ghanaian schools (65 students, 58 lecturers or tutors, 38 supporting staff), the authors build three composites — Guidance and Counselling Administrative Support (GCAS), academic adjustment and intended retention — and fit HC3-robust OLS models, treating the 65 students as the primary analytic population and the full sample as sensitivity evidence. Among students, GCAS predicts adjustment (B = 0.673, p < .001) and, separately, intended retention (B = 0.483, p = .002); in the joint model adjustment remains significant (B = 0.751, p < .001) while GCAS does not (B = -0.022, p = .888), and a 2,000-replication bootstrap gives a positive indirect estimate, 95% percentile interval (0.342, 0.667). The authors conclude that academic adjustment is the most relevant statistical link between counselling-support conditions and intended retention.

Contribution and Significance

The substantive contribution is a role-stratified descriptive account of how perceived administrative provision for school counselling co-varies with adjustment and intended persistence in a Ghanaian sample, together with the observation that adjustment absorbs essentially the whole association. The methodological contribution is a piece of reporting discipline: the authors recognise that student-worded items answered by 96 staff cannot be read as student outcomes, and say so in the Table 3 note itself.

Relative to the closest prior work the advance is small. Gyasi et al. (2025) analysed a cross-sectional Ghanaian student sample linking counselling and psychological services to academic performance, and Panford-Quainoo et al. (2024) documented counselling practices and challenges in Accra senior high schools. What is new here is moving the predictor from service use to perceived administrative conditions — a sensible reframing, but one the manuscript does not establish as measurable, since the GCAS scale is unvalidated and its reliability is α = .55 (Table 1). The largest reported result, the adjustment–retention correlation of r = .730, is also the least surprising.

Strengths

• The decision to make the student-only subsample (n = 65) primary and the mixed-role sample sensitivity evidence is correct and uncommon; the Table 3 note states plainly that those estimates "incorporate 96 staff responses to questions framed as students' personal experience."
• The inferential language is disciplined throughout. The Literature Gap Synthesis states that the indirect estimate "is not an estimate of the causal effect of changing administrative practice," and the manuscript consistently separates intended from observed retention, contrasting itself with Adams (2023).
• Estimation is reproducible as described: HC3 standard errors, 2,000 case-resampled bootstrap replications and a reported seed (20261103).
• The effect decomposition is internally consistent in both samples: 0.673 × 0.751 = 0.505, which with the direct coefficient of -0.022 recovers the reported total of 0.483, and the Table 3 figures reconcile in the same way.

Recommendation

Reject for the Journal of Psychologists and Counsellors in Schools. The evidence does not convincingly support the central claims, and the reason lies in the instrument rather than the analysis, so no further modelling of these data can repair it: the adjustment and retention items name the very support the predictor measures, and two of the ten are verbatim duplicates. The manuscript is also incomplete in ways that prevent assessment — the primary models have no results table, and the sampling frame, response rate, ethics approval and consent procedures are absent. The careful design reporting is genuine and worth preserving, but it is applied to measures that cannot carry the claims.

Major Concerns

1. The adjustment and retention items are contaminated by the predictor, and two items are duplicates (Appendix, SAMPLE QUESTIONS, Sections C and D; Table 1). Three of the five adjustment items attribute the respondent's state to school support: item 11 ("because of support I have received in school") and items 12 and 13 ("after receiving guidance from my teachers or the coordinator"). Retention item 19 does the same ("because of the support available to me"). The association between GCAS and adjustment is therefore partly definitional rather than empirical, which bears directly on the first and third of the paper's three stated propositions.

   Separately, items 12 and 13 are word-for-word identical, as are items 16 and 17, the second of each pair distinguished only by a trailing "2". The Methodology confirms this, listing four content areas for adjustment and four for retention, not five each. A near-tautological item pair inflates Cronbach's alpha, so the α values of .73 and .71 offered as evidence of acceptable reliability overstate it, while the predictor itself stands at α = .55. The authors should report item-level descriptives and inter-item correlations, re-estimate every model on de-duplicated and de-contaminated scales, and state how much of each association survives.

2. The primary results are not reported (APPENDICES; Findings, "Regression Results Among Student Respondents"). The appendices contain Table 1 and Table 3; there is no Table 2. Every student-only estimate appears only in the abstract and in narrative prose, without standard errors, confidence intervals, per-model sample sizes or R²; the statement that "the two variables explained approximately 53% of the differences in students' intended retention" appears in no table. The Findings also assert results that are never evidenced: that students "expressed greater satisfaction with school leadership's encouragement ... than with the availability of private counselling offices and dedicated counselling time," and that coordinator shortages, inadequate facilities and heavy workloads "were the most frequently identified barriers." No item-level means and no frequencies for the Section E checkbox item appear anywhere, yet both claims are carried into the Discussion and the recommendations.

3. Sampling, population and ethics are undocumented (Methodology, "Design, source and analytic population" and "Research ethics and reporting boundaries"; "Funding, ethics and data availability"). The entire account of data collection is that responses were "gathered using google forms": no schools are named or counted, and no recruitment procedure or response rate is given. The population is internally inconsistent — the Introduction refers to "Ghanaian pre-tertiary schools" and the abstract to a single "Ghanaian school as case study," while the questionnaire offers "Lecturer / Tutor" as a role and sets an age floor of "18 - 24 years," although pre-tertiary students in Ghana are predominantly minors. Who the 65 students are is therefore unclear. On ethics the manuscript states only that the dataset is anonymous and defers the rest: "Funding and ethics/consent statements are to be completed from the original research records." For a survey of school students, documented approval and consent (and assent, if minors were involved) are prerequisites for review.

4. The gap is asserted rather than demonstrated, on a six-item reference list. The Literature and conceptual framing section devotes one or two sentences each to five works without establishing what is unknown, and foundational literature on the constructs is absent: no retention model is invoked (Tinto, V., 1993, Leaving College: Rethinking the Causes and Cures of Student Attrition, 2nd ed., University of Chicago Press); no validated adjustment instrument is used or discussed (Baker, R. W., & Siryk, B., 1984, "Measuring adjustment to college," Journal of Counseling Psychology, 31(2), 179–189); the bootstrap procedure is uncited (Preacher, K. J., & Hayes, A. F., 2008, "Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models," Behavior Research Methods, 40(3), 879–891); and the paper's own central caveat has a direct source (Maxwell, S. E., & Cole, D. A., 2007, "Bias in cross-sectional analyses of longitudinal mediation," Psychological Methods, 12(1), 23–44). Without this grounding the claimed gap cannot be evaluated.

Minor Comments

1. With n = 65 and no power consideration, the student-only estimates are imprecise, and the omitted confidence intervals make that invisible.
2. The Findings open the student-focused narrative with M = 2.93, which Table 1 shows is the all-respondent figure, not the student figure of 2.91 used in the abstract.
3. The Acknowledgments, Declaration of interest and Funding sections contain instructions to the author rather than statements, and a stray coding label, "GCAS5.", is left inside questionnaire item 9.
4. The sentence "These findings were consistent with the relationships reported in the original dissertation" (Mixed-Role Sensitivity Analysis) refers to a source document that is neither cited nor described.
5. The prose needs thorough revision for grammar and completeness — for example, "These conditions are cannot fully be justified as desired outcomes in whole," and the incomplete abstract sentence "this study explores their practical implications through a cross-sectional, with Ghanaian school as case study."
6. Because respondents are nested within schools, the models should either account for clustering or the authors should state that all respondents came from one school.
7. The demographic items on sex, age and employment status were collected but no breakdown is reported, and no covariates enter any model.
8. Measurement invariance across roles is never tested, so the mixed-role comparison cannot distinguish genuine agreement from differing reference points.
9. The recommendations in "Implications for Guidance and Counselling and Educational Administration" would follow from the literature without these data, and should be tied to specific reported estimates or softened.