IB Diploma Programme · Research review
Subject difficulty:
General Ability
For students with similar estimated general academic ability, which IB DP subjects are harder to achieve a given grade in?
Read this first
We reproduce Table 4, Dataset 1 from Badham, Meadows and Baird (2025), using historical May 2019 DP results. The authors use a Rasch Partial Credit Model: A student’s chances of higher grades depend on the difference between their estimated general ability and the subject’s grade-specific difficulty.
The model estimates the statistical difficulty for each grade in each subject. Our Overall difficulty is the average of the published estimates at grades 5, 6 and 7. Logits are model units that quantify the level of estimated ability needed to reach a grade. They are not grade points or percentages.
Subjects ranked by difficulty
Overall difficulty · mean of grades 5 to 7. Most difficult first (rank 1); equal values share a rank.
- 1Computer science · SL12.61
- 2Chemistry · HL22.57
- 3Computer science · HL32.52
- 4History · SL42.41
- 5Mathematics · HL52.40
- 6Chemistry · SL62.28
- 7Information technology in a global society · HL72.21
- 8Physics · SL82.11
- 9Mathematics · SL92.11
- 10Physics · HL102.09
- 11English literature · SL112.08
- 12Biology · HL121.95
- 13History · HL131.90
- 14Design technology · HL141.88
- 15Biology · SL151.65
- 16English language & literature · SL161.63
- 16Economics · SL161.63
- 18Environmental systems & societies · SL181.51
- 19English literature · HL191.45
- 20Psychology · HL201.42
- 21Economics · HL211.39
- 22Psychology · SL221.37
- 23Geography · SL231.30
- 24Film · HL241.30
- 25Sports exercise & health science · SL251.24
- 26Philosophy · HL261.23
- 27Philosophy · SL271.17
- 28Spanish literature · SL281.11
- 29Geography · HL290.96
- 30Spanish language & literature · SL300.96
- 31Global politics · HL310.96
- 32Maths studies · SL320.87
- 33Business management · HL330.76
- 34Chinese language & literature · SL340.73
- 35English language & literature · HL350.68
- 36Chinese literature · HL360.61
- 37Spanish literature · HL370.60
- 38Theatre · HL380.59
- 39Business management · SL390.47
- 40Spanish language & literature · HL400.38
- 41Chinese language & literature · HL41-0.14
No subjects match your search.
Method and sources
These are the published Rasch–Thurstone grade thresholds in logits from Dataset 1, linked through “general academic ability”. Higher thresholds indicate greater estimated difficulty at that grade. HL means Higher Level; SL means Standard Level. Historical subject names are retained as published.
Ranks are calculated separately for each measure across all 41 table entries. Equal values share a rank, with subsequent ranks skipped. Searching hides entries without changing ranks. The single marker shows rank, not logit distance; numerical logits are shown alongside. Overall difficulty is our arithmetic mean of the published grade 5, 6 and 7 thresholds, not a separate estimate reported in Table 4. Overall ranks use the unrounded means; displayed logits are rounded to two decimal places. No baseline is calculated.
Source typing error: Table 4 appears to contain a typing error: row 11 is labelled “Philosophy Higher” instead of “Psychology Higher”. We use Psychology HL, consistent with Table 3 and Figure 2. Row 25 remains Philosophy HL. All logits are unchanged; the downloadable CSV retains the original label and correction note.
Dataset 1 excludes subjects and students with poor model fit. The authors emphasize the limits of linking different subjects through a single construct: different linking constructs produce different difficulty orders. Results describe the retained historical sample and model, not all IB DP subjects or an intrinsic measure of subject difficulty. Small differences and rounded ties do not establish statistical significance.
Source: Louise Badham, Michelle Meadows and Jo-Anne Baird (2025), Construct comparability and the limits of post hoc modeling: insights from International Baccalaureate multi-language assessments, Table 4, p. 9. Read the article · Source PDF · Table 4. © 2025 the authors, CC BY. Data extracted; layout and rankings adapted.