Large introductory literature courses that assign Equus often rely on multiple teaching assistants to grade the resulting essays, which introduces a consistency challenge that does not exist when a single instructor grades every paper themselves. Two equally capable TAs can read the same borderline essay on Dysart's moral conflict and land on meaningfully different scores if they have not been trained to apply the rubric with a shared understanding of what each criterion actually means in practice. A structured TA training process, built specifically around the rubric and the particular interpretive challenges of this text, reduces this variation significantly before it ever reaches students.

An effective training session starts by having all TAs independently grade the same two or three sample essays, ideally chosen to represent a range of quality including at least one genuinely borderline case, before comparing scores as a group and discussing where and why disagreements occurred. This process surfaces specific points of confusion, such as differing interpretations of what counts as sufficient engagement with the play's ambiguity around Dysart, in a concrete and immediately relevant way rather than through abstract rubric discussion alone. The discussion that follows these sample gradings is often more valuable than the rubric document itself, since it reveals the specific judgment calls TAs need to align on for this particular text.

Because Equus involves several interpretively tricky areas, moral ambiguity around Dysart, appropriate handling of the play's mature content, and the distinction between religious symbolism identification and interpretation, training should address these text-specific challenges explicitly rather than relying solely on a generic essay-grading rubric. A short written guide supplementing the main rubric, with brief notes on how to handle these specific recurring judgment calls, gives TAs a reference to return to mid-semester when a tricky essay arises and they cannot recall exactly how the group resolved a similar case during initial training. This supplementary guide also becomes useful for future semesters, reducing the training burden each time new TAs join the course.

Building a Shared Anchor Set

Beyond the initial training session, maintaining a shared set of anchor essays, papers previously agreed upon as representing specific score bands, gives TAs an ongoing reference point throughout the grading period rather than relying purely on memory of the initial calibration discussion. When a TA encounters an essay that feels difficult to place on the scale, comparing it directly against an anchor essay from a similar score band often resolves the uncertainty faster than re-reading the full rubric language in isolation. Keeping this anchor set updated across semesters, adding new examples as particularly clear or particularly ambiguous cases arise, builds an increasingly useful training resource over time.

• Has all TAs independently grade the same sample essays before comparing and discussing scores
• Creates a supplementary guide addressing text-specific judgment calls unique to Equus
• Maintains an ongoing anchor set of essays representing each score band for reference
• Schedules a brief mid-semester check-in to recalibrate before grading drifts apart
• Documents specific disagreements from training discussions for future reference

Spotting and Correcting Drift Mid-Semester

Even after strong initial training, grading standards can drift apart over the course of a semester as each TA independently reads dozens of essays and gradually develops their own individual sense of what counts as strong or weak analysis, often without realizing the drift is happening. A brief mid-semester check-in, where TAs again independently grade a shared sample essay and compare results, catches this drift while there is still time to correct it before the bulk of the semester's grading is complete. This kind of periodic recalibration matters more for a text as interpretively complex as Equus than it would for a more straightforward assignment, since the room for reasonable disagreement is naturally wider.

When drift is identified, the most useful response is usually a short group discussion of the specific essay that revealed the disagreement, walking through exactly why different TAs landed on different scores and what the correct or agreed-upon interpretation should be going forward. Simply telling TAs to "grade more consistently" without pointing to a concrete example rarely produces real behavior change, whereas a specific, shared case study tends to stick and genuinely realign subsequent grading. Documenting the resolution of this discussion, adding it to the supplementary guide mentioned earlier, also prevents the same disagreement from resurfacing later in the semester.

Using Shared Digital Rubrics to Support Consistency

Digital grading platforms that apply a shared, detailed rubric consistently across every TA's grading can provide an additional layer of consistency beyond training sessions alone, since the platform applies the exact same criteria and scoring logic regardless of which individual TA is reviewing a given essay. This does not eliminate the need for human judgment on the more nuanced interpretive questions, but it does provide a consistent baseline that human graders can check their own scoring against, flagging cases where a TA's score diverges significantly from what the rubric-based assessment would suggest. For a course with several TAs each grading their own section, this kind of built-in consistency check can catch drift much earlier than a periodic manual check-in alone.

Course instructors who oversee TA grading also benefit from being able to review aggregate patterns across all sections at once, seeing at a glance whether one TA's section is scoring systematically higher or lower than others on the same assignment, which is much harder to notice by spot-checking individual essays alone. Tools that surface this kind of cross-section comparison give instructors an early warning system for grading inconsistency, letting them intervene with targeted feedback to a specific TA before a whole semester's worth of grades reflects an uncorrected drift. For a course built around a text as interpretively rich and occasionally tricky as Equus, that kind of oversight capability is often what actually keeps grading fair across an entire large course.