Supplemental Material


Return to main supplement page

Individual observer's data: GTH (Experiment 3)


The tables below provide individual observer data for the blocks task, as tab delimited text.
For each level of scene complexity, we provide one data file for all conditions (illumination-constant, illuminant-changed with local contrast preserved and illuminant-change with local contrast silenced combined).



Block selection data (Simple scene, Complex scene)
Data tables follow this format: 
Column 1: Condition code (NC for illuminant-constant; CB for bluish illuminant-change with local contrast preserved; CS for bluish illuminant-change with local contrast silenced).
Column 2: Block number
Column 3: Trial type (denoted as a unique integer; refers to a specific combination of target, competitor pair and illumination condition)
Column 4: Target 1 "Straw": f
irst competitor in the competitor pair on a given trial  (Convention for illumination-constant condition: 1= C-2, 2 =  C-1; 3 = T; 4 = C+1; 5 = C+2; For illumination-changed condition 1= T, 2 = C1; 3 = C2; 4 = C3; 5 = R; See Figure 2).
Column 5: Target 1 "Straw": second competitor in the competitor pair on a given trial (same convention as in Column 4).
Column 6: Target 1 "Straw": competitor chosen by the subject on a given trial (1 = the first competitor in a pair, 2 = the second competitor in a pair).
Column 7: Target 2 "Skin": first competitor in the competitor pair on a given trial  (same convention as in Column 4).
Column 8: Target 2 "Skin": second competitor in the competitor pair on a given trial (same convention as in Column 4).
Column 9: Target 2 "Skin": competitor chosen by the subject on a given trial (1 = the first competitor in a pair, 2 = the second competitor in a pair).
Column 10: Target 3 "Lettuce": first competitor in the competitor pair on a given trial  (same convention as in Column 4).
Column 11: Target 3 "Lettuce": second competitor in the competitor pair on a given trial (same convention as in Column 4).
Column 12: Target 3 "Lettuce": competitor chosen by the subject on a given trial (1 = the first competitor in a pair, 2 = the second competitor in a pair).
Column 13: Target 4 "Bark": first competitor in the competitor pair on a given trial  (same convention as in Column 4).
Column 14: Target 4 "Bark": second competitor in the competitor pair on a given trial (same convention as in Column 4).
Column 15: Target 4 "Bark": competitor chosen by the subject on a given trial (1 = the first competitor in a pair, 2 = the second competitor in a pair).
Column 16: Reaction time in seconds (from the trial onset until the subject made a choice).
Column 17: Session order number.
Note: NaN value for competitor chosen (in Columns 6, 9, 12 or 15) indicates that no block from the source was assigned to this target block (and the corresponding block in the workspace remained black) or that the block that was assigned was not one of the two competitors for this target (and was thus excluded from the analysis).

Figures showing subject's choices matrices and inferred matches for each target and conditon.
The data is represented in the format of a heat map, hence slightly different than in Figure 3A.
For each pairwise comparison of competitors the matrix shows the proportion of time one competitor (vertical axis) is chosen relative to the other (horizontal axis).
The entries above and below the diagonal in the matrices are not independent: entries above the diagonal are obtained as 1 minus the corresponding entry below the diagonal. Cells along the diagonal were set to 0.5, since we did not have subjects select between two identical tests.
The circles in each matrix show the position of the subject’s selection-bagth matches inferred from these responses. The circle color is intended to illustrate (only roughly) the target color.
Simple scene condition: Illuminant-constant, Illuminant-changed (contrast preserved), Illuminant-changed (contrast silenced).
Complex scene condition: Illuminant-constant, Illuminant-changed (contrast preserved), Illuminant-changed (contrast silenced).

Figure showing the relation between measured selection probabilities and those predicted via MLDS procedure
Simple scene (pdf), Complex scene (pdf) 
Different illumination condition is plotted in each row: top: illuminant-constant (NC); middle: illuminant-change to bluish with local contrast preserved (CB); bottom: illuminant-change to bluish with local contrast silenced (CS).
Each target is plotted in a different column (marked as T1 to T4). Each subplot follows the same conventions as Figure 3B.