Skip to main content

Letter Logic

Letter Logic is a constrained word-guessing task that measures vocabulary breadth, hypothesis testing, constraint satisfaction, and working memory updating. Participants guess a secret 5-letter word, receiving color-coded feedback after each attempt: green for correct position, yellow for correct letter in wrong position, and gray for letters not in the word. The task combines linguistic knowledge with logical deduction.

ActivitySpec: lamp.letter_logic

Cognitive domain: Vocabulary, hypothesis testing, constraint satisfaction, working memory

Configuration​

SettingDescription
DifficultyControls number of rounds and guesses per round
API settings fields
Dashboard SettingAPI FieldTypeDefault
Difficultydifficultystring"medium"

Sample Instructions​

"Guess the secret 5-letter word in 6 tries. After each guess, tiles change color: green means the letter is correct and in the right spot, yellow means the letter is in the word but in the wrong spot, and gray means the letter is not in the word."

Usage​

Target words are drawn from a pool of 2,003 common 5-letter words (SUBTLEX-US Zipf frequency ≥ 3.5), and valid guesses come from a broader list of 5,567 words (Zipf ≥ 2.0). Letter evaluation uses a two-pass system: exact position matches (green) are identified first, then wrong-position matches (yellow).

An on-screen QWERTY keyboard displays color-coded keys tracking letter status across guesses, helping the participant track which letters have been confirmed, relocated, or eliminated.

Difficulty levels control the structure: Easy (3 rounds, 8 guesses each), Medium (5 rounds, 6 guesses each), and Hard (7 rounds, 5 guesses each).

Scoring​

Primary metrics are rounds solved, mean guesses to solve, and guess efficiency (proportion of guess budget used). The guess distribution shows frequency of solving in 1, 2, 3, etc. guesses.

References
  1. Berger, J. (2023). Wordle and constraint satisfaction: cognitive mechanisms underlying word-guessing games. Cognitive Science, 47(3), e13281.

Data​

static_data​

FieldTypeDescription
rounds_solvednumberRounds where the word was found
rounds_failednumberRounds where guesses ran out
rounds_totalnumberTotal rounds played
mean_guesses_to_solvenumberAverage guesses needed on solved rounds
mean_guess_efficiencynumberProportion of guess budget used
mean_solve_time_msnumberAverage time to solve a round (ms)
guess_distributionnumber[]Frequency of solving in 1, 2, 3... guesses
difficultystringDifficulty level used
max_guessesnumberGuess budget per round
scorenumberOverall score
correct_answersnumberRounds solved
total_questionsnumberTotal rounds
questionnaireobjectPost-game ratings: clarity (1-5), happiness (1-5)

temporal_slices​

One entry per round.

FieldTypeDescription
itemnumberRound index
typestringRound type
round_numbernumberRound sequence number
target_wordstringThe secret word
solvedbooleanWhether the word was found
guesses_usednumberNumber of guesses made
max_guessesnumberGuess budget for this round
guessesstring[]All guesses in order
rt_msnumberTotal round time (ms)
durationnumberTotal round duration (ms)

Cortex Features​

No Cortex features are currently available for this activity.

View in Portal | Python SDK | API Reference