Rejected from Computers and Education? The 7 Best Journals to Submit Next

The fastest, cheapest cascade after Computers & Education stays inside the Elsevier portfolio first, then steps out by rejection reason.

If the editor offered a transfer, start there. Computers & Education submits through Elsevier Editorial Manager, which supports the Article Transfer Service. When an editor decides your work fits a sister title better, you get an emailed transfer offer; accepting it moves the manuscript and any reviewer comments to the recommended journal without a new submission. For AI-in-education papers, that recommended title is usually Computers & Education: Artificial Intelligence.

Letting the reviews travel with the paper is the single biggest time saver available to you, so do not decline a transfer just to re-pick the journal yourself.

If it was a desk rejection for scope (no transfer offered), step down by fit, not by tier. A paper desk-rejected as too narrow for the broad Computers & Education readership is not necessarily a worse paper. It is a fit problem. Route an AI-tutoring or learning-analytics study to Computers & Education: Artificial Intelligence; route a design-research or instructional-design study to Educational Technology Research and Development; route a post-secondary online-learning study to Internet and Higher Education; route a single-platform deployment to Journal of Computer Assisted Learning or Education and Information Technologies.

If it was a post-review rejection, fix the named gap before you move. Reviewer concerns about a missing learning-outcome measure, a single-context design, or a thin theory contribution will reappear at every serious edtech journal. The next tier down does not lower that bar; it only changes the masthead. Address the substance, then submit. Many of these journals will accept a brief response explaining how you used the prior reviews.

The ladder, in one line: transfer offer first, then Computers & Education: Artificial Intelligence or British Journal of Educational Technology as the next best targets, then Educational Technology Research and Development / Internet and Higher Education, then Journal of Computer Assisted Learning / Education and Information Technologies as the realistic next venue for narrower work.

Map the rejection reason to the next venue with this routing table:

Source: publisher aims-and-scope pages for each venue, accessed June 2026. Routing logic is Manusights editorial judgment based on stated journal scope.

Recent Computers & Education articles you can check your own scope against include 10.1016/j.compedu.2026.105563, 10.1016/j.compedu.2026.105593, and 10.1016/j.compedu.2026.105592 on ScienceDirect; if your study reads narrower than those, the routing table above is the realistic next step.

In our pre-submission review work with Computers and Education submissions, three patterns generate the most consistent desk rejections and post-review declines, and each is testable against your own manuscript before you resubmit anywhere. (Per Elsevier published medians, Computers & Education issues a first decision in about 7 days and a post-review decision in about 48 days, so the desk screen below is where most papers are stopped.)

The outcome battery has no validated learning-outcome measure. This is the pattern we see most often in Computers & Education manuscripts we review. Authors run a study on one tutoring system, VR module, or classroom platform and report satisfaction scales, intention-to-use items, self-reported perceived learning, time-on-task, or click-stream engagement, then stop.

Computers & Education editors specifically check whether the outcome set includes at least one validated learning measure: a pre/post conceptual-knowledge test with reported reliability, a validated educational-psychology construct, a delayed-retention or transfer task, or a real performance result such as grade or pass rate. If your statistical analysis reports only perceptual or behavioral-intent outcomes, expect a desk redirect.

The fix is to add one validated learning-outcome measure, report effect sizes with confidence intervals rather than bare p-values, and lead the abstract with the learning result.

The AI architecture is the contribution instead of the learning change. Computers & Education manuscripts built on generative AI, intelligent tutoring, or automated feedback frequently lead with the model novelty (a fine-tuning approach, a retrieval pipeline, a prompting scheme) and treat the educational consequence as a downstream demo.

Computers & Education treats that inversion as a venue mismatch: the published criterion is that the technology must change learning, teaching, feedback, self-regulation, or assessment, not that it is novel in computer-science terms. When the contribution statement names a model or architecture as the primary advance, the manuscript gets flagged at desk.

The fix touches two manuscript components at once: rewrite the contribution statement so the educational change is the load-bearing claim, demote the architecture detail to the methods or supplementary section, and ground the design in a learning theory that predicts the effect.

The single-context deployment has no generalizability or theory contribution. The third pattern is a well-run study at one institution, one course, or one cohort that never tells readers outside that setting what to take from it.

Computers & Education reviewers check the discussion for explicit boundary conditions (which conditions and learner populations the findings should and should not transfer to), a named theory contribution (which framework the results extend or refine, not just cite), and design implications stated as patterns rather than as local recommendations. A manuscript that reads as a parochial case study draws a revision request for these additions or a redirect to a narrower venue.

The fix is to add a generalizability paragraph naming two or three contexts where the design should and should not transfer, state one concrete way the theory moves forward, and own the single-context limitation honestly rather than burying it.

These are the same checks a Computers and Education submission readiness check runs against your manuscript before you commit to the next journal.

Do not just blast the same file down the ladder. A Computers & Education rejection is a free, expert read on where your manuscript is weak, and the next journal will see the same gaps if you ignore it.

Separate the two rejection types honestly. A desk rejection for scope or a missing learning-outcomes connection is mostly a routing problem: pick the better-fit journal above and reframe the contribution, and you can move quickly. A post-review rejection that named a missing validated learning measure, a single-context design, or a thin theory contribution is a substance problem: the same reviewers' concerns will surface at British Journal of Educational Technology, Educational Technology Research and Development, or anywhere else serious. That needs real work, not a new masthead.

Two more honest checks. First, confirm the new journal's article types and formatting match your paper, because Education and Information Technologies, for example, no longer accepts review papers, and each venue has its own abstract cap (Computers & Education uses a 250 word abstract; others differ), so submitting the wrong type or an over-length package wastes a cycle.

Second, if your study is a single-context deployment, decide now whether you can add the generalizability and theory paragraphs that every reputable edtech journal expects; if you cannot, target a narrower venue deliberately rather than collecting a second rejection at a broad one.