How Does the (Re)Presentation of Instructions Influence Their Implementation?

Cai S. Longman; Baptist Liefooghe; Frederick Verbruggen

doi:10.5334/joc.63

Figures & Tables

Miniblock instructions from each experiment (upper panel) and the timeline of a single trial during the NEXT and GO phases (lower panel) of each miniblock. In all experiments, the block number and miniblock number were presented at the top of the instructions screen which remained visible until the participant pressed the space bar (minimum 3 seconds). The additional instruction presented in Experiment 2 informed the participant which effector to respond with during the GO phase (little fingers in the example).

NEXT RTs (top), Correct NEXT RTs (middle) and NEXT PEs (bottom) from Experiments 1–2. Error bars show the between-subjects standard error of the mean difference between compatible and incompatible trials in each condition.

Table 1

Omnibus ANOVA Results from Experiments 1a and 1b. Equivalent Bayes Factors are also Reported.

	NEXT RT
Effect	DF	MSE	F	p	BF	η²
Experiment 1a
Instructions	(1,39)	9,583.56	8.23	0.007	0.135 ± 3.3%	0.067
Compatibility	(1,39)	12,383.29	42.85	<0.001	<0.001 ± 3.3%	0.324
Compatibility × Instructions	(1,39)	6,357.96	4.05	0.051	1.368 ± 3.8%	0.023
Experiment 1b
Abstractness	(1,39)	15,808.33	1.18	0.284	3.663 + 4.4%	0.008
Compatibility	(1,39)	26,505.30	6.53	0.015	0.092 + 5.6%	0.073
Compatibility × Abstractness	(1,39)	13,948.24	0.51	0.481	3.565 + 4.0%	0.003
	Correct NEXT RT
Effect	DF	*MSE*	F	p	BF	η²
Experiment 1a
Instructions	(1,39)	9,434.66	7.94	0.008	0.145 ± 5.8%	0.065
Compatibility	(1,39)	11,586.41	29.61	<0.001	<0.001 ± 6.7%	0.241
Compatibility × Instructions	(1,39)	6,713.58	4.19	0.047	1.089 ± 5.9%	0.025
Experiment 1b
Abstractness	(1,39)	16,549.46	0.70	0.407	4.294 + 3.9%	0.005
Compatibility	(1,39)	24,205.70	5.17	0.029	0.247 + 5.5%	0.056
Compatibility × Abstractness	(1,39)	13,718.95	0.22	0.640	3.824 + 3.8%	0.001
	NEXT PE
Effect	DF	MSE	F	p	BF	η²
Experiment 1a
Instructions	(1,39)	15.08	1.00	0.322	4.552 ± 3.8%	0.006
Compatibility	(1,39)	37.00	16.25	<0.001	<0.001 ± 4.1%	0.190
Compatibility × Instructions	(1,39)	13.78	0.96	0.333	3.533 ± 2.7%	0.005
Experiment 1b
Abstractness	(1,39)	9.07	1.99	0.166	1.981 + 4.0%	0.018
Compatibility	(1,39)	7.78	8.79	0.005	0.100 + 4.1%	0.063
Compatibility × Abstractness	(1,39)	9.07	1.99	0.166	1.570 + 3.9%	0.018

[i] Note: Bayes factors indicate whether removal of the effect/interaction from the model would materially impair its fit. Thus, Bayes factors <1 indicate that the effect/interaction is an important contributor to the model.

Table 2

T-Test Results from Experiment 1a. Equivalent Bayes factors are also reported.

	NEXT RT
	Difference	lower CI	upper CI	DF	t	p	BF	g_av
Exemplar-based	141 ms	99	182	39	6.86	<0.001	414,283.458	0.834
Label-based	90 ms	44	136	39	3.95	<0.001	84.625	0.477
	Correct NEXT RT
	Difference	lower CI	upper CI	DF	t	p	BF	g_av
Exemplar-based	119 ms	76	163	39	5.54	<0.001	7,945.345	0.698
Label-based	66 ms	23	109	39	3.11	0.004	10.064	0.355
	NEXT PE
	Difference	lower CI	upper CI	DF	t	p	BF	g_av
Exemplar-based	4.5%	2	7	39	3.49	0.001	25.815	0.897
Label-based	3.3%	1	5	39	3.45	0.001	23.513	0.902

[i] Note: p-values in bold font survived Holms-Bonferroni correction for multiple comparisons.

Table 3

Descriptive Statistics for GO Performance in all Experiments. Standard Errors and 95% Confidence Intervals are also Reported.

		GO RT (ms)				GO PE (%)
Experiment	Condition	Mean	SE	lower CI	upper CI	Mean	SE	lower CI	upper CI
Exp 1a	Exemplar-based	735	27	680	790	5.3	0.6	4.2	6.5
	Label-based	689	22	644	734	5.0	0.6	3.9	6.1
Exp 1b		875	38	798	953	9.1	0.8	7.4	10.8
Exp 2	Same fingers	568	15	538	599	8.1	0.7	6.7	9.5
	Different fingers	579	15	549	609	8.7	0.8	7.1	10.4
	Feet	630	18	594	665	7.0	0.6	5.7	8.3

Table 4

Omnibus ANOVA Results from Experiment 2. Equivalent Bayes Factors are also Reported.

	NEXT RT
Effect	DF	MSE	F	p	BF	η²
Effector	(2,94)	4380.65	1.478	0.233	6.364 + 5.4%	0.013
Compatibility	(1,47)	6236.84	16.28	<0.001	<0.001 + 4.0%	0.096
Effector*Compatibility	(2,94)	2681.63	11.5	<0.001	0.023 + 4.5%	0.061
	Correct NEXT RT
Effect	DF	*MSE*	F	p	BF	η²
Effector	(2,94)	3516.26	1.484	0.232	5.213 + 2.2%	0.015
Compatibility	(1,47)	3779.39	19.755	<0.001	<0.001 + 2.7%	0.1
Effector*Compatibility	(2,94)	1705.54	8.371	<0.001	0.215 + 2.3%	0.041
	NEXT PE
Effect	DF	*MSE*	F	p	BF	η²
Effector	(2,94)	9.81	2.558	0.083	5.002 + 2.6%	0.015
Compatibility	(1,47)	23.32	4.074	0.049	0.295 + 2.9%	0.028
Effector*Compatibility	(2,94)	13.43	10.446	<0.001	0.002 + 2.7%	0.079

[i] Note: Bayes factors indicate whether removal of the effect/interaction from the model would materially impair its fit. Thus, Bayes factors <1 indicate that the effect/interaction is an important contributor to the model.

Table 5

T-Test Results from Experiment 2. Equivalent Bayes Factors are also Reported.

	NEXT RT
Effector	Difference	lower CI	upper CI	DF	t	p	BF	g_av
Same fingers	79 ms	47	111	47	5.03	<0.001	2,467.881	0.569
Different fingers	18 ms	–3	39	47	1.68	0.099	0.580	0.131
Feet	16 ms	–7	38	47	1.43	0.161	0.403	0.117
	Correct NEXT RT
Effector	Difference	lower CI	upper CI	DF	t	p	BF	g_av
Same fingers	60 ms	37	83	47	5.25	<0.001	4,970.890	0.476
Different fingers	17 ms	–1	34	47	1.87	0.067	0.782	0.131
Feet	20 ms	1	39	47	2.09	0.042	1.141	0.153
	NEXT PE
Effector	Difference	lower CI	upper CI	DF	t	p	BF	g_av
Same fingers	3.9%	1.5	6.3	47	3.30	0.002	16.823	0.751
Different fingers	–0.4%	–1.8	1.0	47	–0.63	0.534	0.189	0.112
Feet	0.0%	–0.9	0.8	47	–0.11	0.915	0.158	0.016

[i] Note: p-values in bold font survived Holms-Bonferroni correction for multiple comparisons.

Table 6

Results from the Correlations between the Absolute⁶ NEXT Compatibility Effect and GO performance in all Experiments.

Experiment	Condition	RT			PE
		DF	r	p	DF	r	p
Exp 1a	Exemplar-based	39	0.07	0.672	39	0.10	0.520
	Label-based	39	0.04	0.797	39	0.20	0.210
Exp 1b		39	0.07	0.656	39	0.02	0.886
Exp 2	Same finger	47	0.03	0.826	47	0.12	0.400
	Different finger	47	–0.04	0.797	47	–0.19	0.190
	Feet	47	–0.16	0.288	47	0.11	0.466

[i] Note: We used correct NEXT RTs for the latency analyses.

Table A1

Descriptive Statistics for Experiment 2. Standard Errors and 95% Confidence Intervals are also Reported.

			NEXT RT (ms)
Half	Effector	Compatibility	Mean	SE	lower CI	upper CI
1st	Same fingers	Compatible	606	19	568	645
		Incompatible	707	22	663	752
	Different fingers	Compatible	643	19	604	681
		Incompatible	673	20	632	713
	Feet	Compatible	653	23	607	700
		Incompatible	675	19	637	713
2nd	Same fingers	Compatible	522	20	481	562
		Incompatible	579	23	533	625
	Different fingers	Compatible	578	21	535	620
		Incompatible	582	23	536	628
	Feet	Compatible	552	22	508	596
		Incompatible	563	20	522	603
			Correct NEXT RT (ms)
			Mean	SE	lower CI	upper CI
1st	Same fingers	Compatible	600	19	562	639
		Incompatible	675	20	635	715
	Different fingers	Compatible	626	18	589	662
		Incompatible	652	19	614	690
	Feet	Compatible	634	23	588	680
		Incompatible	657	19	619	695
2nd	Same fingers	Compatible	507	18	471	544
		Incompatible	555	20	515	596
	Different fingers	Compatible	555	19	516	593
		Incompatible	561	21	519	603
	Feet	Compatible	530	18	494	567
		Incompatible	546	19	508	584
			NEXT PE (%)
			Mean	SE	lower CI	upper CI
1st	Same fingers	Compatible	0.7	0.3	0.1	1.3
		Incompatible	5.5	1.2	3.1	7.9
	Different fingers	Compatible	3.2	0.6	2.0	4.4
		Incompatible	3.9	0.8	2.3	5.6
	Feet	Compatible	2.3	0.5	1.3	3.3
		Incompatible	2.2	0.5	1.2	3.2
2nd	Same fingers	Compatible	1.2	0.6	0.0	2.4
		Incompatible	4.3	1.2	1.9	6.7
	Different fingers	Compatible	3.8	0.9	1.9	5.6
		Incompatible	2.1	0.6	0.9	3.3
	Feet	Compatible	2.2	0.6	1.0	3.5
		Incompatible	2.4	0.6	1.1	3.6

Table A2

Omnibus ANOVA Results from Experiment 2. Equivalent Bayes Factors are also Reported.

	NEXT RT
Effect	DF	MSE	F	p	BF	η²
Half	(1,47)	13,939.84	97.20	<0.001	<0.001 ± 6.6%	0.283
Effector	(2,94)	8,771.23	1.26	0.289	9.760 ± 6.9%	0.006
Compatibility	(1,47)	12,551.15	16.12	<0.001	<0.001 ± 6.7%	0.056
Half × Effector	(2,94)	3,817.78	3.53	0.033	4.150 ± 7.4%	0.008
Half × Compatibility	(1,47)	3,274.07	7.75	0.008	1.359 ± 12.5%	0.007
Effector × Compatibility	(2,94)	5,347.06	11.74	<0.001	0.004 ± 7.0%	0.035
Half × Effector × Compatibility	(2,94)	3,731.67	0.85	0.430	9.070 ± 6.7%	0.002
	Correct NEXT RT
Effect	DF	*MSE*	F	p	BF	η²
Half	(1,47)	12,384.41	112.23	<0.001	<0.001 ± 22.6%	0.347
Effector	(2,94)	6,904.48	1.32	0.272	7.025 ± 22.5%	0.007
Compatibility	(1,47)	7,327.39	20.53	<0.001	<0.001 ± 22.4%	0.054
Half × Effector	(2,94)	3,462.49	3.15	0.047	3.067 ± 22.7%	0.008
Half × Compatibility	(1,47)	2,766.86	4.28	0.044	1.951 ± 22.6%	0.004
Effector × Compatibility	(2,94)	3,304.81	9.27	<0.001	0.104 ± 26.7%	0.023
Half × Effector × Compatibility	(2,94)	2,969.63	0.38	0.686	10.022 ± 22.9%	0.001
	NEXT PE
Effect	DF	*MSE*	F	p	BF	η²
Half	(1,47)	23.43	0.60	0.444	8.260 ± 5.4%	0.001
Effector	(2,94)	19.88	2.44	0.092	5.652 ± 10.8%	0.009
Compatibility	(1,47)	46.14	4.17	0.047	0.134 ± 6.5%	0.017
Half × Effector	(2,94)	12.60	0.49	0.612	22.484 ± 6.0%	0.001
Half × Compatibility	(1,47)	13.77	4.22	0.046	2.087 ± 3.9%	0.005
Effector × Compatibility	(2,94)	27.89	10.02	<0.001	<0.001 ± 43.0%	0.048
Half × Effector × Compatibility	(2,94)	15.73	1.33	0.270	6.295 ± 4.4%	0.004

[i] Note: Bayes factors indicate whether removal of the effect/interaction from the model would materially impair its fit. Thus, Bayes factors <1 indicate that the effect/interaction is an important contributor to the model.

How Does the (Re)Presentation of Instructions Influence Their Implementation?

Figures & Tables

Figure 1

Figure 2

Table 1

Table 2

Table 3

Table 4

Table 5

Table 6

Table A1

Table A2

Paradigm

My account