Academia.eduAcademia.edu

Figure 9 – uploaded by Christiane Lange-Kuettner

See full PDFdownloadDownload figure
Figure 9. Correlations between area, array exploration and place sequencing (N = 522). (A) Human players (n = 261). (B) Machine (n = 261).

Figure 9 Correlations between area, array exploration and place sequencing (N = 522). (A) Human players (n = 261). (B) Machine (n = 261).

Related Figures (13)

Figure 1. Coloring-in of a 20 x 20 grid. Note. green = human (dark grey); yel- low = machine (light grey); black = empty.
Figure 1. Coloring-in of a 20 x 20 grid. Note. green = human (dark grey); yel- low = machine (light grey); black = empty.
Figure 2. Graphic representation and data generation of coloring-in grid moves. (A) Human player. (B) Machine player. Note. An area was evaluated as places that had at least one shared side with another place of the same color. Places touching at corners did not count as areas. In each figure, the blue (online only) squares denote the largest area created and the grey areas are the smaller clusters created by the respective player. The white places represent the areas and places colored in by the opponent. In this example, the human player created a cohesive large area of 158 places and the machine player created a large area of 73 places.
Figure 2. Graphic representation and data generation of coloring-in grid moves. (A) Human player. (B) Machine player. Note. An area was evaluated as places that had at least one shared side with another place of the same color. Places touching at corners did not count as areas. In each figure, the blue (online only) squares denote the largest area created and the grey areas are the smaller clusters created by the respective player. The white places represent the areas and places colored in by the opponent. In this example, the human player created a cohesive large area of 158 places and the machine player created a large area of 73 places.
Figure 3. Average time to color a place per instruction (n = 261, human players only). Note. The error bars denote the standard deviation. Horizontal bars indi- cate significant differences (please see the text for details). A 8 (cluster) x 4 (instruction) mixed ANOVA with repeated measures of cluster size was conducted. The main effect of instruction, F(3, 261) = 7.50, p< 0.001, Np = 0.08, was sig- nificant. Pairwise comparisons within the model showed that all instructed conditions differed significantly from the control condition, p, < 0.010, but not from each other, p, = 1.0. All instructed conditions, with either time pressure or area instruction, yielded the same overall effect of the aver- age area being about one to two places larger than in the
Figure 3. Average time to color a place per instruction (n = 261, human players only). Note. The error bars denote the standard deviation. Horizontal bars indi- cate significant differences (please see the text for details). A 8 (cluster) x 4 (instruction) mixed ANOVA with repeated measures of cluster size was conducted. The main effect of instruction, F(3, 261) = 7.50, p< 0.001, Np = 0.08, was sig- nificant. Pairwise comparisons within the model showed that all instructed conditions differed significantly from the control condition, p, < 0.010, but not from each other, p, = 1.0. All instructed conditions, with either time pressure or area instruction, yielded the same overall effect of the aver- age area being about one to two places larger than in the
Figure 4. Cluster size (n = 261, human players only). Note. The error bars denote the standard deviation. All clusters differed significantly from each other, see the statistical values in Table A1. In Figure 5, all special instructions appear to yield a larger large cluster than the standard instruction. However, post-hoc univariate analyses, see Table A2, showed that for most cluster sizes, as hypothesized, only the area instruc- tions (area and area/speed) yielded clusters that were signifi- cantly different to those of the control group. For the largest cluster, the explicit instruction to build an area was indeed conducive to build a larger area with a difference of about thirty places. Thereafter, the high spatial constraints induced by the large area caused the instruction to become ineffect- ive for the second largest area (see Cluster 2 in Table A2). Moreover, for the smaller clusters, the explicit area instruc- tions had the opposite effect, namely, the area instruction groups built smaller small areas than without such explicit instruction.
Figure 4. Cluster size (n = 261, human players only). Note. The error bars denote the standard deviation. All clusters differed significantly from each other, see the statistical values in Table A1. In Figure 5, all special instructions appear to yield a larger large cluster than the standard instruction. However, post-hoc univariate analyses, see Table A2, showed that for most cluster sizes, as hypothesized, only the area instruc- tions (area and area/speed) yielded clusters that were signifi- cantly different to those of the control group. For the largest cluster, the explicit instruction to build an area was indeed conducive to build a larger area with a difference of about thirty places. Thereafter, the high spatial constraints induced by the large area caused the instruction to become ineffect- ive for the second largest area (see Cluster 2 in Table A2). Moreover, for the smaller clusters, the explicit area instruc- tions had the opposite effect, namely, the area instruction groups built smaller small areas than without such explicit instruction.
Figure 5. Two-way interaction of instruction and cluster size (n = 261, human players only). Note. The error bars denote the standard error
Figure 5. Two-way interaction of instruction and cluster size (n = 261, human players only). Note. The error bars denote the standard error
Figure 6. Average sequence length per instruction (n = 261, human players only). Note. The error bars denote the standard deviation. The horizontal bar indicates a significant difference (please see the text for details).
Figure 6. Average sequence length per instruction (n = 261, human players only). Note. The error bars denote the standard deviation. The horizontal bar indicates a significant difference (please see the text for details).
Figure 7. Two-way interaction of cluster size in humans and the machine (N = 522). Note. The bars denote the standard error.
Figure 7. Two-way interaction of cluster size in humans and the machine (N = 522). Note. The bars denote the standard error.
Note. “Adjustment for multiple comparisons: Bonferroni. Table 1. Post-hoc t-tests (two-tailed, equal variances not assumed): human vs. machine cluster size (N = 522).
Note. “Adjustment for multiple comparisons: Bonferroni. Table 1. Post-hoc t-tests (two-tailed, equal variances not assumed): human vs. machine cluster size (N = 522).
Figure 8. Spatial relationships between places in human and machine (N = 522). Note. The bars denote the standard error.
Figure 8. Spatial relationships between places in human and machine (N = 522). Note. The bars denote the standard error.
Note. Based on estimated marginal means. *Adjustment for multiple comparisons: Bonferroni. Table A1. Post-hoc tests of the cluster size (humans only, N = 261). Appendix
Note. Based on estimated marginal means. *Adjustment for multiple comparisons: Bonferroni. Table A1. Post-hoc tests of the cluster size (humans only, N = 261). Appendix
Note. “Adjustment for multiple pairwise comparisons: Bonferroni. Instruction abbreviations are Control=C, Speed=S, Area=A, Area/Speed = AS. Significan comparisons are printed in bold. Table A2. Post-hoc ANOVAs of the instruction effect on cluster size (n = 261) followed by pairwise comparisons within the model.
Note. “Adjustment for multiple pairwise comparisons: Bonferroni. Instruction abbreviations are Control=C, Speed=S, Area=A, Area/Speed = AS. Significan comparisons are printed in bold. Table A2. Post-hoc ANOVAs of the instruction effect on cluster size (n = 261) followed by pairwise comparisons within the model.
Note. Based on estimated marginal means. “Adjustment for multiple comparisons: Bonferroni. Table A3. Post-hoc tests of the cluster size (humans/machine, N = 522).
Note. Based on estimated marginal means. “Adjustment for multiple comparisons: Bonferroni. Table A3. Post-hoc tests of the cluster size (humans/machine, N = 522).
Note. Correlations for human participants are above the diagonal, for the machine system below the diagonal. **Correlation is significant at the 0.01 level (2. tailed). *Correlation is significant at the 0.05 level (2-tailed). Only Bonferroni-adjusted correlations are reported in the text. Table A4. Pearson correlations (two-tailed) between colored place distribution variables and area building (N = 522).
Note. Correlations for human participants are above the diagonal, for the machine system below the diagonal. **Correlation is significant at the 0.01 level (2. tailed). *Correlation is significant at the 0.05 level (2-tailed). Only Bonferroni-adjusted correlations are reported in the text. Table A4. Pearson correlations (two-tailed) between colored place distribution variables and area building (N = 522).
Related topics:
Computer ScienceGrid ComputingHuman Computer InteractionGame Theory (Psychology)Space and PlaceSpatial Cognition (Psychology)
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved