Seeing Spelling Isn’t Enough


Alisha Tabassum 1

Shazia Ali1 Helene Slaattelid Øya1 Sofia Tsitsopoulou1 Jens Roeser1

1 Nottingham Institute of Psychology, Nottingham Trent University

Introduction

  • Writers must decide what to mention first and when to begin typing while retrieving spellings for upcoming words (Flower & Hayes, 1980; Roeser et al., 2019).
  • Spelling difficulty may affect written sentence production before the difficult word is typed, during the word, or downstream in later sentence regions.
  • We ask whether spelling difficulty affects word order or emerges during real-time production.

Finding

  • Spelling difficulty did not change horizontal word order, but it did alter the typing time course.
  • Seeing a spelling immediately before writing was not enough to remove hesitation for high-diversity words.
  • The strongest evidence is therefore in post-onset production dynamics, not in the initial ordering decision.

Methods

Task

Sentence production task with spelling prompts.

Schematic trial flow.

Figure 1: Schematic trial flow for ‘The piranha is next to the swing’.

Design

Spelling diversity (high vs low) was manipulated.

Table 1: Spelling-diversity calculation for one high-diversity target item. Separate picture-naming sample: 97 participants named this picture; 74 used the target name.
Spelling Frequency Proportion Spelling diversity
colander 30 40.5% 2.44
collander 16 21.6%
collinder 13 17.6%
colinder 7 9.5%
calender 1 1.4%
caulinder 1 1.4%
colandor 1 1.4%
colender 1 1.4%
colindar 1 1.4%
colonder 1 1.4%
colunder 1 1.4%
culinder 1 1.4%
  • Spelling-prompt alignment was manipulated with correct, plausible, and incorrect prompts.
  • Object pictures originally came from the Rossion and Pourtois colourized Snodgrass and Vanderwart set (Rossion & Pourtois, 2004); images were cleaned, standardized, and improved using AI.
  • Spatial arrangements placed at 3, 9, 12, or 6 o’clock, creating horizontal is next to trials and vertical is above/is below trials; horizontal trials allowed left-to-right / right-to-left descriptions.
Table 2: Example item sets crossing spelling diversity, spelling prompt, and other object.
Item High-diversity High prompts Low-diversity Low prompts Other object
1 piranha correct: piranha
plausible: pirahna
incorrect: prianha 		jellyfish correct: jellyfish
plausible: jelyfish
incorrect: jellyfihs 		swing
2 pomegranate correct: pomegranate
plausible: pomegranite
incorrect: poemgranate 		watermelon correct: watermelon
plausible: watermelan
incorrect: waetrmelon 		saw
3 meringue correct: meringue
plausible: merangue
incorrect: merignue 		biscuit correct: biscuit
plausible: biscuitt
incorrect: bsicuit 		watch
4 colander correct: colander
plausible: collander
incorrect: colanedr 		bagpipes correct: bagpipes
plausible: bagpieps
incorrect: bapgipes 		pear
  • 48 item sets and 24 filler pairs (diagonal pairs with unrelated prime) per list.
  • Counterbalancing crossed spelling diversity, spelling prompt, and target position within item.

Participants

  • Broad analysis: 194 participants.
  • Time-course model: 144 participants and 5,034 trials.
  • Age: 34.3 years on average (SD = 8.4, range 18-50).

Results

Spelling Decision

Plausible spellings were accepted much more often for high-diversity than low-diversity targets. Aligned means correct or plausible prompts accepted.

Spelling-prompt acceptance and spelling-decision time by spelling diversity.

Figure 2: Spelling-prompt acceptance and spelling-decision time by spelling diversity.

Horizontal Starting Direction

  • Horizontal next to trials showed a near-ceiling spatial left-to-right bias (99.6-100.0%; 95% PI lower bounds at least 98.6%), leaving little room for spelling difficulty to affect starting direction.
  • Primed-picture starts simply tracked where the prime appeared: 99.97%-99.99% when the prime was left, but only 0.24%-0.45% when it was right.
  • In next to filler trials, left-to-right starts were 98.6% (97.4%-99.3%) when the left image was upper-left and 66.0% (54.0%-76.6%) when it was lower-left.

Hesitation Probability

Hesitation probabilities were determined using Bayesian mixture modelling (Roeser et al., 2024, 2025).

High-minus-low diversity hesitation probability from paired posterior contrasts across horizontal and vertical trials.

Figure 3: High-minus-low diversity hesitation probability from paired posterior contrasts across horizontal and vertical trials.

References

Flower, L. S., & Hayes, J. R. (1980). The dynamics of composing: Making plans and juggling constraints. In L. W. Gregg & E. R. Steinberg (Eds.), Cognitive processes in writing: An interdisciplinary approach (pp. 31–50). Lawrence Erlbaum.
Roeser, J., Conijn, R., Chukharev, E., Ofstad, G. H., & Torrance, M. (2025). Typing in tandem: Language planning in multisentence text production is fundamentally parallel. Journal of Experimental Psychology: General, 154(7), 1824–1854. https://doi.org/10.1037/xge0001759
Roeser, J., De Maeyer, S., Leijten, M., & Van Waes, L. (2024). Modelling typing disfluencies as finite mixture process. Reading and Writing, 37(2), 359–384. https://doi.org/10.1007/s11145-021-10203-z
Roeser, J., Torrance, M., & Baguley, T. (2019). Advance planning in written and spoken sentence production. Journal of Experimental Psychology: Learning, Memory, and Cognition, 45(11), 1983–2009. https://doi.org/10.1037/xlm0000685
Rossion, B., & Pourtois, G. (2004). Revisiting Snodgrass and Vanderwart’s object pictorial set: The role of surface detail in basic-level object recognition. Perception, 33(2), 217–236. https://doi.org/10.1068/p5117

Spelling difficulty persists into the time course of sentence production.