Effects of Cognitive Load and Emotional Load on Memory Selectivity - An Individual Differences Perspective

MD & BJS

2022-06-27

Online Data

Method

Participants

133 participants completed the study. We excluded 30 participants with low accuracy (< 70% correct) in the study phase and 11 with low accuracy in the test phase (hit - false alarm rate < 0). The final sample consisted of 61 women and 31 men. Mean age was 19 years (SD = 1).

check if four versions are counterbalanced across participants

For study phase:

expName	n
V1_part1	24
V2_part1	24
V3_part1	20
V4_part1	24

For test phase:

expName	n
V1_part2	24
V2_part2	24
V3_part2	20
V4_part2	24

Material

Procedure

In the study phase, participants completed a task switching procedure was administered.
As a filler task and to assess WMC, participants completed the letter-memory task.
In the test phase, a surprise recognition memory test assessed participant’s memory for stimuli presented in the study phase
Motivation measures: AATQ, BIS/BAS, Self-Control Scale (in random order)

Results

Study Phase

Descriptives

transition	M_rt	SD_rt	SE_rt	M_acc	SD_acc	SE_acc
repeat	1153	356	19	0.913	0.127	0.007
switch	2009	593	31	0.897	0.133	0.007

plot accuracy

plot response times

Inference

T-test on accuracy:

## 
##  Paired t-test
## 
## data:  acc by transition
## t = 2.8662, df = 91, p-value = 0.00516
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  0.004796261 0.026453739
## sample estimates:
## mean of the differences 
##                0.015625

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	602.808	2.074	26444.330	0.000	*	0.988
transition	1	91	0.045	0.498	8.215	0.005	*	0.006
emotionality	1	91	0.799	0.660	110.131	0.000	*	0.098
task	1	91	3.296	2.174	137.947	0.000	*	0.310
transition:emotionality	1	91	0.013	0.467	2.516	0.116		0.002
transition:task	1	91	0.000	0.490	0.044	0.835		0.000
emotionality:task	1	91	0.960	0.551	158.536	0.000	*	0.116
transition:emotionality:task	1	91	0.006	0.429	1.251	0.266		0.001

Main effect of emotionality is significant, but no interaction with transition

dfstudy %>%
  group_by(subject, emotionality) %>%
  summarise(acc = mean(acc)) %>%
  group_by(emotionality) %>%
  summarise(M = mean(acc), SE = se(acc), n = length(acc))

## `summarise()` has grouped output by 'subject'. You can override using the
## `.groups` argument.

## # A tibble: 2 x 4
##   emotionality     M      SE     n
##   <fct>        <dbl>   <dbl> <int>
## 1 negative     0.872 0.00638    92
## 2 neutral      0.938 0.00640    92

dfstudy %>%
  group_by(subject, task) %>%
  summarise(acc = mean(acc)) %>%
  group_by(task) %>%
  summarise(M = mean(acc), SE = se(acc), n = length(acc))

## `summarise()` has grouped output by 'subject'. You can override using the
## `.groups` argument.

## # A tibble: 2 x 4
##   task        M      SE     n
##   <fct>   <dbl>   <dbl> <int>
## 1 picture 0.972 0.00621    92
## 2 word    0.838 0.00939    92

The word task was more difficult (not surprising)

dfstudy %>%
  group_by(subject, task, emotionality) %>%
  summarise(acc = mean(acc)) %>%
  group_by(task, emotionality) %>%
  summarise(M = mean(acc), SE = se(acc), n = length(acc))

## `summarise()` has grouped output by 'subject', 'task'. You can override using
## the `.groups` argument.
## `summarise()` has grouped output by 'task'. You can override using the
## `.groups` argument.

## # A tibble: 4 x 5
## # Groups:   task [2]
##   task    emotionality     M      SE     n
##   <fct>   <fct>        <dbl>   <dbl> <int>
## 1 picture negative     0.975 0.00608    92
## 2 picture neutral      0.969 0.00704    92
## 3 word    negative     0.769 0.0116     92
## 4 word    neutral      0.907 0.0103     92

A negative word made it even more difficult!

T-test on reaction times:

## 
##  Paired t-test
## 
## data:  rt by transition
## t = -23.061, df = 91, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -929.1784 -781.8007
## sample estimates:
## mean of the differences 
##               -855.4895

The effects of task switching on performance (accuracy and reaction times) was replicated.

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	1.840103e+09	82320993	2034.103	0.000	*	0.918
transition	1	91	1.346627e+08	23043016	531.801	0.000	*	0.449
task	1	91	7.761008e+06	15787515	44.735	0.000	*	0.045
emotionality	1	91	4.989384e+05	8297416	5.472	0.022	*	0.003
transition:task	1	91	1.892525e+06	13039179	13.208	0.000	*	0.011
transition:emotionality	1	91	2.005931e+04	8244497	0.221	0.639		0.000
task:emotionality	1	91	5.977360e+04	7996865	0.680	0.412		0.000
transition:task:emotionality	1	91	3.765931e+03	6538056	0.052	0.819		0.000

Does the interaction between task switching and task reflect asymmetric switch costs? Remember: Picture task = easy, word task = hard

dfstudy %>%
  group_by(subject, task, transition) %>%
  summarise(rt = mean(rt)) %>%
  group_by(task, transition) %>%
  summarise(M = mean(rt), SE = se(rt), n = length(rt))

## `summarise()` has grouped output by 'subject', 'task'. You can override using
## the `.groups` argument.
## `summarise()` has grouped output by 'task'. You can override using the
## `.groups` argument.

## # A tibble: 4 x 5
## # Groups:   task [2]
##   task    transition     M    SE     n
##   <fct>   <fct>      <dbl> <dbl> <int>
## 1 picture repeat     1000.  22.6    92
## 2 picture switch     1957.  54.9    92
## 3 word    repeat     1307.  35.9    92
## 4 word    switch     2061.  56.6    92

Yes! Switch costs are greater for the picture task, which is the easy task. Switching away from the harder task is more difficult.

If switching away from the harder task poses a greater cognitive load, this should also be reflected in memory selectivity. –> greater switch costs to memory selectivity for the pictures. BUT: Pictures are generally better remembered than words. This makes it unreasonable to test this hypothesis with the present design.

Test phase

We analyze only the Picture Recognition test because emotionality was induced by having negative and neutral words and we do not want to have a confound of material. Pictures were always neutral and they were counterbalanced across conditions and participants (words not).

Descriptives

Hits and False Alarms

Overall the hit rate was 0.562 (SE = 0.015) and the false alarms rate was 0.131 (SE = 0.013).

The first step is to replicate the task switching x attention interaction on recognition performance

Means per condition:

## `summarise()` has grouped output by 'attention'. You can override using the
## `.groups` argument.

attention	transition	m_acc	sd_acc	se_acc
target	repeat	0.763	0.193	0.014
target	switch	0.745	0.192	0.014
distractor	repeat	0.337	0.201	0.015
distractor	switch	0.404	0.207	0.015

plot recognition performance

ANOVA results:

## Warning: Collapsing data to cell means. *IF* the requested effects are a subset
## of the full design, you must use the "within_full" argument, else results may be
## inaccurate.

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	116.297	7.536	1404.332	0.00	*	0.908
attention	1	91	13.490	2.525	486.173	0.00	*	0.535
transition	1	91	0.056	0.746	6.841	0.01	*	0.005
attention:transition	1	91	0.169	0.916	16.740	0.00	*	0.014

The interaction between attention and transition is critical here. The significant interaction replicates previous studies and means that task switching reduces memory selectivity.

This can also be seen in the main effect of transition on memory selectivity (Hits Targets - Hits Distractors).

Plot memory selectivity

This plot shows that memory selectivity is higher for repeat vs. switch trials. This is the task switching effect on memory selectivity.

This effect should be based on remember responses (vs. know responses)

ANOVA for remember:

## Warning: Collapsing data to cell means. *IF* the requested effects are a subset
## of the full design, you must use the "within_full" argument, else results may be
## inaccurate.

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	33.381	11.599	261.895	0.000	*	0.646
attention	1	91	4.699	5.378	79.515	0.000	*	0.204
transition	1	91	0.007	0.656	0.944	0.334		0.000
attention:transition	1	91	0.073	0.657	10.053	0.002	*	0.004

## Warning: Collapsing data to cell means. *IF* the requested effects are a subset
## of the full design, you must use the "within_full" argument, else results may be
## inaccurate.

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	25.065	14.617	156.051	0.000	*	0.544
attention	1	91	2.266	4.925	41.864	0.000	*	0.097
transition	1	91	0.024	0.781	2.770	0.099		0.001
attention:transition	1	91	0.020	0.711	2.553	0.114		0.001

The interaction between task switching and attention was only significant for remember responses, not for know responses. This replicates previous research

Next, we introduce the factor Emotional Load to check the main effect on memory selectivity and a potential interaction with Cognitive Load (transition)

Please remember that we are looking only at memory for pictures (which were paired either with a emotional word or neutral word)

Before running the analysis for the full design (attention x transition x emotionality) let’s first check the interaction of emotional load with attention on recognition performance (hits)

Emotionality denotes if the word (which was presented over the picture) was negative or neutral

Compared to the cognitive load effect the pattern goes in the opposite direction! –> higher memory selectivity with emotional load

Emotional load leads to higher memory selectivity. While

Cognitive load leads to lower memory selectivity.

Let’s run the analysis for the full design. Again first for recognition performance and then for memory selectivity (which is the same just for better understanding of the meaning I show both)

2 x 2 x 2 ANOVA (attention x emotionality x transition)

Cognitive Load effect:

For Targets memory is better for repeat than switch

for Distractors memory is better for switch than repeat

Emotional Load effect:

For Targets memory is better if the Distractor was negative than neutral

For Distractors memory is better if the Target was neutral than negative

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	232.594	15.072	1404.332	0.000	*	0.891
attention	1	91	26.980	5.050	486.173	0.000	*	0.486
emotionality	1	91	0.025	0.999	2.236	0.138		0.001
transition	1	91	0.112	1.491	6.841	0.010	*	0.004
attention:emotionality	1	91	0.182	1.386	11.967	0.001	*	0.006
attention:transition	1	91	0.337	1.832	16.740	0.000	*	0.012
emotionality:transition	1	91	0.002	1.459	0.132	0.717		0.000
attention:emotionality:transition	1	91	0.000	1.263	0.001	0.979		0.000

Memory selectivity

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	53.961	10.100	486.173	0.000	*	0.739
emotionality	1	91	0.365	2.773	11.967	0.001	*	0.019
transition	1	91	0.674	3.664	16.740	0.000	*	0.034
emotionality:transition	1	91	0.000	2.526	0.001	0.979		0.000

There is no significant interaction between cognitive load and emotional load.

As cognitive load and emotional load affect memory selectivity in opposite ways and don’t interact suggests that there are different mechanisms going on.

Check if there is a dissociation between remember and know responses.

For remember:

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	66.761	23.197	261.895	0.000	*	0.618
attention	1	91	9.398	10.755	79.515	0.000	*	0.185
emotionality	1	91	0.041	1.164	3.214	0.076		0.001
transition	1	91	0.014	1.313	0.944	0.334		0.000
attention:emotionality	1	91	0.030	1.189	2.264	0.136		0.001
attention:transition	1	91	0.145	1.313	10.053	0.002	*	0.003
emotionality:transition	1	91	0.012	1.262	0.882	0.350		0.000
attention:emotionality:transition	1	91	0.000	1.156	0.027	0.870		0.000

For know:

Effect	DFn	DFd	SSn	SSd	F	p	p<.05	ges
(Intercept)	1	91	50.130	29.233	156.051	0.000	*	0.520
attention	1	91	4.531	9.850	41.864	0.000	*	0.089
emotionality	1	91	0.002	1.060	0.182	0.670		0.000
transition	1	91	0.048	1.563	2.770	0.099		0.001
attention:emotionality	1	91	0.065	1.042	5.678	0.019	*	0.001
attention:transition	1	91	0.040	1.421	2.553	0.114		0.001
emotionality:transition	1	91	0.004	0.755	0.501	0.481		0.000
attention:emotionality:transition	1	91	0.000	1.318	0.016	0.899		0.000

The attention x emotionality interaction was only significant for know responses, not for remember responses. This is very interesting as it supports the interpretation that the effects of task switching and emotionality on memory selectivity stem from different mechanisms.

The next step is to introduce the individual differences variables.

Individual Differences

Descriptives

For WMC:

Here are the descriptives for the raw and transformed data

N	M	SD	Min	Max	Skewness	Kurtosis	Measure
92	0.88	0.12	0.50	1.00	-1.16	3.85	PropCorrect
92	2.86	0.15	2.33	3.00	-1.28	4.06	M correct Letters
92	0.00	1.00	-3.10	0.95	-1.16	3.85	PropCorrect (z-stand.)
92	1.29	0.23	0.79	1.57	-0.14	2.01	PropCorrect (arcsine trans.)

Miyake et al. (2000) report positive skew, after arcsine transformation: .35

Histogram raw data

It’s strongly skewed.

Histogram arcsine transformed data

Still not perfect :( Is this good enough?

For future studies we should use the more difficult Letter-Memory task. (recall 4 letters instead of 3, number of letters presented: 5,7,9, or 11)

For BIS/BAS:

Here are the descriptives for the raw and transformed data

N	M	SD	Min	Max	Skewness	Kurtosis	Measure
92	20.62	4.00	7.00	28.00	-0.73	3.91	BIS
92	0.00	1.00	-3.41	1.85	-0.73	3.91	BIS (z-stand.)
92	37.38	5.94	20.00	50.00	-0.07	3.32	BAS
92	0.00	1.00	-2.93	2.12	-0.07	3.32	BAS (z-stand.)

Histogram raw data

Histogram transformed data

For Approach/Avoidance:

Here are the descriptives for the raw and transformed data

N	M	SD	Min	Max	Skewness	Kurtosis	Measure
92	4.25	1.34	1.00	6.33	-0.55	2.47	Avoidance
92	0.00	1.00	-2.42	1.55	-0.55	2.47	Avoidance (z-stand.)
92	4.98	0.90	2.83	6.67	-0.09	2.53	Approach
92	0.00	1.00	-2.38	1.88	-0.09	2.53	Approach (z-stand.)

Histograms raw data

For Selfcontrol:

Here are the descriptives for the raw and transformed data

N	M	SD	Min	Max	Skewness	Kurtosis	Measure
92	40.78	8.29	20.00	56.00	-0.26	2.72	Selfcontrol
92	0.00	1.00	-2.51	1.84	-0.26	2.72	Selfcontrol (z-stand.)

Histogram raw data

save(old_long, file = "./Datafiles/old_long.Rda")
rm(list = ls())
load("./Datafiles/old_long.Rda")

Working memory capacity (WMC)

WMC was measured with the letter memory task.

ANCOVA hits

pes = partial eta squared ges = general eta squared

## Warning: Numerical variables NOT centered on 0 (i.e., likely bogus results):
## d.propCorrect

Effect	df	MSE	F	pes	p.value
d.propCorrect	1, 90	0.17	0.46	.005	.498
attention	1, 90	0.05	1.12	.012	.292
d.propCorrect:attention	1, 90	0.05	4.21 *	.045	.043
transition	1, 90	0.02	0.56	.006	.457
d.propCorrect:transition	1, 90	0.02	0.15	.002	.697
emotionality	1, 90	0.01	1.59	.017	.211
d.propCorrect:emotionality	1, 90	0.01	1.13	.012	.290
attention:transition	1, 90	0.02	0.26	.003	.610
d.propCorrect:attention:transition	1, 90	0.02	1.18	.013	.280
attention:emotionality	1, 90	0.02	1.31	.014	.256
d.propCorrect:attention:emotionality	1, 90	0.02	0.46	.005	.502
transition:emotionality	1, 90	0.02	0.00	<.001	.962
d.propCorrect:transition:emotionality	1, 90	0.02	0.01	<.001	.922
attention:transition:emotionality	1, 90	0.01	0.36	.004	.552
d.propCorrect:attention:transition:emotionality	1, 90	0.01	0.37	.004	.546

Same analysis with centered variable:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 90	0.17	0.46	.005	.498
attention	1, 90	0.05	503.32 ***	.848	<.001
propCorrect.z:attention	1, 90	0.05	4.21 *	.045	.043
transition	1, 90	0.02	6.78 *	.070	.011
propCorrect.z:transition	1, 90	0.02	0.15	.002	.697
emotionality	1, 90	0.01	2.24	.024	.138
propCorrect.z:emotionality	1, 90	0.01	1.13	.012	.290
attention:transition	1, 90	0.02	16.77 ***	.157	<.001
propCorrect.z:attention:transition	1, 90	0.02	1.18	.013	.280
attention:emotionality	1, 90	0.02	11.90 ***	.117	<.001
propCorrect.z:attention:emotionality	1, 90	0.02	0.46	.005	.502
transition:emotionality	1, 90	0.02	0.13	.001	.718
propCorrect.z:transition:emotionality	1, 90	0.02	0.01	<.001	.922
attention:transition:emotionality	1, 90	0.01	0.00	<.001	.979
propCorrect.z:attention:transition:emotionality	1, 90	0.01	0.37	.004	.546

ANCOVA remember responses:

## Warning: Numerical variables NOT centered on 0 (i.e., likely bogus results):
## d.propCorrect

Effect	df	MSE	F	pes	p.value
d.propCorrect	1, 90	0.26	0.03	<.001	.871
attention	1, 90	0.12	0.00	<.001	.950
d.propCorrect:attention	1, 90	0.12	1.72	.019	.194
transition	1, 90	0.01	2.56	.028	.113
d.propCorrect:transition	1, 90	0.01	2.19	.024	.143
emotionality	1, 90	0.01	0.12	.001	.731
d.propCorrect:emotionality	1, 90	0.01	0.01	<.001	.921
attention:transition	1, 90	0.01	2.61	.028	.110
d.propCorrect:attention:transition	1, 90	0.01	4.33 *	.046	.040
attention:emotionality	1, 90	0.01	0.01	<.001	.929
d.propCorrect:attention:emotionality	1, 90	0.01	0.01	<.001	.906
transition:emotionality	1, 90	0.01	0.62	.007	.432
d.propCorrect:transition:emotionality	1, 90	0.01	0.86	.009	.356
attention:transition:emotionality	1, 90	0.01	1.52	.017	.221
d.propCorrect:attention:transition:emotionality	1, 90	0.01	1.61	.018	.208

Same analysis with centered variable

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 90	0.26	0.03	<.001	.871
attention	1, 90	0.12	80.14 ***	.471	<.001
propCorrect.z:attention	1, 90	0.12	1.72	.019	.194
transition	1, 90	0.01	0.96	.011	.331
propCorrect.z:transition	1, 90	0.01	2.19	.024	.143
emotionality	1, 90	0.01	3.18 +	.034	.078
propCorrect.z:emotionality	1, 90	0.01	0.01	<.001	.921
attention:transition	1, 90	0.01	10.42 **	.104	.002
propCorrect.z:attention:transition	1, 90	0.01	4.33 *	.046	.040
attention:emotionality	1, 90	0.01	2.24	.024	.138
propCorrect.z:attention:emotionality	1, 90	0.01	0.01	<.001	.906
transition:emotionality	1, 90	0.01	0.88	.010	.351
propCorrect.z:transition:emotionality	1, 90	0.01	0.86	.009	.356
attention:transition:emotionality	1, 90	0.01	0.03	<.001	.870
propCorrect.z:attention:transition:emotionality	1, 90	0.01	1.61	.018	.208

ANCOVA know responses

## Warning: Numerical variables NOT centered on 0 (i.e., likely bogus results):
## d.propCorrect

Effect	df	MSE	F	pes	p.value
d.propCorrect	1, 90	0.32	0.40	.004	.528
attention	1, 90	0.11	0.65	.007	.421
d.propCorrect:attention	1, 90	0.11	0.01	<.001	.936
transition	1, 90	0.02	0.52	.006	.471
d.propCorrect:transition	1, 90	0.02	0.93	.010	.339
emotionality	1, 90	0.01	0.74	.008	.393
d.propCorrect:emotionality	1, 90	0.01	0.86	.009	.357
attention:transition	1, 90	0.02	0.89	.010	.349
d.propCorrect:attention:transition	1, 90	0.02	0.53	.006	.468
attention:emotionality	1, 90	0.01	1.50	.016	.223
d.propCorrect:attention:emotionality	1, 90	0.01	0.82	.009	.367
transition:emotionality	1, 90	0.01	0.91	.010	.342
d.propCorrect:transition:emotionality	1, 90	0.01	1.13	.012	.291
attention:transition:emotionality	1, 90	0.01	0.32	.004	.575
d.propCorrect:attention:transition:emotionality	1, 90	0.01	0.34	.004	.559

Same analysis with centered variable:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 90	0.32	0.40	.004	.528
attention	1, 90	0.11	41.41 ***	.315	<.001
propCorrect.z:attention	1, 90	0.11	0.01	<.001	.936
transition	1, 90	0.02	2.77 +	.030	.100
propCorrect.z:transition	1, 90	0.02	0.93	.010	.339
emotionality	1, 90	0.01	0.18	.002	.671
propCorrect.z:emotionality	1, 90	0.01	0.86	.009	.357
attention:transition	1, 90	0.02	2.54	.027	.115
propCorrect.z:attention:transition	1, 90	0.02	0.53	.006	.468
attention:emotionality	1, 90	0.01	5.67 *	.059	.019
propCorrect.z:attention:emotionality	1, 90	0.01	0.82	.009	.367
transition:emotionality	1, 90	0.01	0.50	.006	.480
propCorrect.z:transition:emotionality	1, 90	0.01	1.13	.012	.291
attention:transition:emotionality	1, 90	0.01	0.02	<.001	.899
propCorrect.z:attention:transition:emotionality	1, 90	0.01	0.34	.004	.559

Same conclusions with the centered variable.

The interaction WMC x attention x transition was only significant for remember responses, not for know responses. Makes sense as the interaction attention x transition is based on remember responses.

See scatterplots:

for WMC x attention (= main effect of WMC on memory selectivity):

For recognition performance:

## `geom_smooth()` using formula 'y ~ x'

high WMC individuals remember not more but more selectively!

For remember responses:

## `geom_smooth()` using formula 'y ~ x'

for know:

## `geom_smooth()` using formula 'y ~ x'

For remember responses check interaction WMC x attention x transition

## `geom_smooth()` using formula 'y ~ x'
## `geom_smooth()` using formula 'y ~ x'

Nice but both variables are skewed. p_remember is right skewed d.propCorrect is left skewed

old_long$p_remember.arcsine <- asin(sqrt(old_long$p_remember))

p <- ggplot(old_long, aes(x=propCorrect.arcsine, y=p_remember.arcsine, color = attention:transition)) + 
  geom_point() +
  geom_smooth(method=lm, se=TRUE, fullrange=TRUE, level=0.95, alpha = 0.2) +
  theme_classic()

ggMarginal(p, type = "histogram", groupColour = TRUE, groupFill = TRUE)

## `geom_smooth()` using formula 'y ~ x'
## `geom_smooth()` using formula 'y ~ x'

Even after the transformation the distribution does not look great. It looks more like we have 3 groups.

For simplicity let’s use a median split for the analysis. (disabled)

plot

2x2x2x2 ANOVA: - Attention (target vs. distractor) - Transition (switch vs. repeat) - WMC (higher vs. lower) - Emotionality (negative, neutral)

2x2x2 ANOVA on memory selectivity

With the median split analysis we find a significant interaction between WMC and transition (result number 6). This is in line with the theory and hypotheses.

Follow-up analyses

high WMC group

lower WMC group

The task switching effect on memory selectivity is only significant for Higher WMC participants. Why? Need to check if the groups differ in number of errors in study phase. There are also much more participants in the “higher WMC” group (n = 58) than in the lower WMC group (n = 34) –> could be a factor

Before going deeper into the individual differences variables, I want to check if the in person data replicates these findings. Especially the effect of emotional load needs to be replicated because it has the opposite effect on memory selectivity than cognitive load.

Covariate BIS & BAS

As we use negative words, the BIS is most relevant here because this system is relevant for negative affect. For control purposes we also have a look at BAS

For hits:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.17	0.57	.006	.452
BIS.z	1, 89	0.17	0.54	.006	.464
attention	1, 89	0.05	508.73 ***	.851	<.001
propCorrect.z:attention	1, 89	0.05	4.87 *	.052	.030
BIS.z:attention	1, 89	0.05	1.97	.022	.164
transition	1, 89	0.02	6.84 *	.071	.010
propCorrect.z:transition	1, 89	0.02	0.29	.003	.589
BIS.z:transition	1, 89	0.02	1.85	.020	.178
emotionality	1, 89	0.01	2.26	.025	.136
propCorrect.z:emotionality	1, 89	0.01	0.82	.009	.368
BIS.z:emotionality	1, 89	0.01	2.04	.022	.157
attention:transition	1, 89	0.02	16.68 ***	.158	<.001
propCorrect.z:attention:transition	1, 89	0.02	1.00	.011	.321
BIS.z:attention:transition	1, 89	0.02	0.52	.006	.473
attention:emotionality	1, 89	0.02	11.78 ***	.117	<.001
propCorrect.z:attention:emotionality	1, 89	0.02	0.51	.006	.479
BIS.z:attention:emotionality	1, 89	0.02	0.16	.002	.694
transition:emotionality	1, 89	0.02	0.13	.001	.720
propCorrect.z:transition:emotionality	1, 89	0.02	0.01	<.001	.911
BIS.z:transition:emotionality	1, 89	0.02	0.02	<.001	.896
attention:transition:emotionality	1, 89	0.01	0.00	<.001	.979
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.36	.004	.548
BIS.z:attention:transition:emotionality	1, 89	0.01	0.00	<.001	.976

For remember responses

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.26	0.04	<.001	.841
BIS.z	1, 89	0.26	0.13	.001	.718
attention	1, 89	0.12	80.49 ***	.475	<.001
propCorrect.z:attention	1, 89	0.12	2.06	.023	.155
BIS.z:attention	1, 89	0.12	1.39	.015	.242
transition	1, 89	0.01	0.96	.011	.330
propCorrect.z:transition	1, 89	0.01	2.54	.028	.115
BIS.z:transition	1, 89	0.01	1.21	.013	.274
emotionality	1, 89	0.01	3.15 +	.034	.079
propCorrect.z:emotionality	1, 89	0.01	0.02	<.001	.883
BIS.z:emotionality	1, 89	0.01	0.19	.002	.662
attention:transition	1, 89	0.01	10.35 **	.104	.002
propCorrect.z:attention:transition	1, 89	0.01	3.96 *	.043	.050
BIS.z:attention:transition	1, 89	0.01	0.39	.004	.533
attention:emotionality	1, 89	0.01	2.22	.024	.140
propCorrect.z:attention:emotionality	1, 89	0.01	0.02	<.001	.898
BIS.z:attention:emotionality	1, 89	0.01	0.01	<.001	.919
transition:emotionality	1, 89	0.01	0.87	.010	.353
propCorrect.z:transition:emotionality	1, 89	0.01	0.74	.008	.393
BIS.z:transition:emotionality	1, 89	0.01	0.29	.003	.589
attention:transition:emotionality	1, 89	0.01	0.03	<.001	.870
propCorrect.z:attention:transition:emotionality	1, 89	0.01	1.33	.015	.253
BIS.z:attention:transition:emotionality	1, 89	0.01	0.95	.011	.332

For know:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.32	0.52	.006	.471
BIS.z	1, 89	0.32	0.73	.008	.396
attention	1, 89	0.11	40.97 ***	.315	<.001
propCorrect.z:attention	1, 89	0.11	0.00	<.001	.958
BIS.z:attention	1, 89	0.11	0.06	<.001	.811
transition	1, 89	0.02	2.74	.030	.101
propCorrect.z:transition	1, 89	0.02	0.84	.009	.363
BIS.z:transition	1, 89	0.02	0.10	.001	.747
emotionality	1, 89	0.01	0.19	.002	.666
propCorrect.z:emotionality	1, 89	0.01	0.53	.006	.470
BIS.z:emotionality	1, 89	0.01	3.48 +	.038	.065
attention:transition	1, 89	0.02	2.51	.027	.116
propCorrect.z:attention:transition	1, 89	0.02	0.56	.006	.458
BIS.z:attention:transition	1, 89	0.02	0.05	<.001	.823
attention:emotionality	1, 89	0.01	5.62 *	.059	.020
propCorrect.z:attention:emotionality	1, 89	0.01	0.92	.010	.339
BIS.z:attention:emotionality	1, 89	0.01	0.32	.004	.572
transition:emotionality	1, 89	0.01	0.50	.006	.481
propCorrect.z:transition:emotionality	1, 89	0.01	0.91	.010	.342
BIS.z:transition:emotionality	1, 89	0.01	0.79	.009	.377
attention:transition:emotionality	1, 89	0.01	0.02	<.001	.899
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.23	.003	.634
BIS.z:attention:transition:emotionality	1, 89	0.01	0.88	.010	.351

BIS x emotionality just missed significance :(

Now let’s see how BIS relates to proportion know responses for emotional and neutral trials.

Although the BISxemotionality interaction was non significant (p = .06) I want to see the scatter plot because an interaction was hypothesized.

## `geom_smooth()` using formula 'y ~ x'

For neutral trials we have a pretty much straight line. Makes sense, as BIS scores should not affect memory for neutral trials.

For negative trials, however, we have a negative correlation between recognition performance and BIS.

For remember:

## `geom_smooth()` using formula 'y ~ x'

What about the BAS?

For know:

## `geom_smooth()` using formula 'y ~ x'

For remember:

## `geom_smooth()` using formula 'y ~ x'

Regression lines go in opposite directions for remember and know responses. Good!

Covariate Avoidance

For hits:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.17	0.62	.007	.432
avoidance.z	1, 89	0.17	0.53	.006	.467
attention	1, 89	0.05	500.64 ***	.849	<.001
propCorrect.z:attention	1, 89	0.05	4.57 *	.049	.035
avoidance.z:attention	1, 89	0.05	0.52	.006	.472
transition	1, 89	0.02	6.71 *	.070	.011
propCorrect.z:transition	1, 89	0.02	0.18	.002	.671
avoidance.z:transition	1, 89	0.02	0.07	<.001	.798
emotionality	1, 89	0.01	2.28	.025	.134
propCorrect.z:emotionality	1, 89	0.01	0.61	.007	.435
avoidance.z:emotionality	1, 89	0.01	2.76	.030	.100
attention:transition	1, 89	0.02	16.59 ***	.157	<.001
propCorrect.z:attention:transition	1, 89	0.02	1.17	.013	.282
avoidance.z:attention:transition	1, 89	0.02	0.01	<.001	.924
attention:emotionality	1, 89	0.02	11.78 ***	.117	<.001
propCorrect.z:attention:emotionality	1, 89	0.02	0.37	.004	.547
avoidance.z:attention:emotionality	1, 89	0.02	0.12	.001	.729
transition:emotionality	1, 89	0.02	0.13	.001	.719
propCorrect.z:transition:emotionality	1, 89	0.02	0.03	<.001	.862
avoidance.z:transition:emotionality	1, 89	0.02	0.22	.002	.643
attention:transition:emotionality	1, 89	0.01	0.00	<.001	.979
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.18	.002	.673
avoidance.z:attention:transition:emotionality	1, 89	0.01	1.11	.012	.295

For remember responses

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.26	0.07	<.001	.787
avoidance.z	1, 89	0.26	0.44	.005	.507
attention	1, 89	0.12	80.17 ***	.474	<.001
propCorrect.z:attention	1, 89	0.12	2.13	.023	.148
avoidance.z:attention	1, 89	0.12	1.03	.011	.312
transition	1, 89	0.01	0.95	.011	.332
propCorrect.z:transition	1, 89	0.01	1.76	.019	.188
avoidance.z:transition	1, 89	0.01	0.58	.006	.448
emotionality	1, 89	0.01	3.14 +	.034	.080
propCorrect.z:emotionality	1, 89	0.01	0.01	<.001	.932
avoidance.z:emotionality	1, 89	0.01	0.01	<.001	.940
attention:transition	1, 89	0.01	10.31 **	.104	.002
propCorrect.z:attention:transition	1, 89	0.01	4.10 *	.044	.046
avoidance.z:attention:transition	1, 89	0.01	0.01	<.001	.928
attention:emotionality	1, 89	0.01	2.22	.024	.139
propCorrect.z:attention:emotionality	1, 89	0.01	0.05	<.001	.827
avoidance.z:attention:emotionality	1, 89	0.01	0.38	.004	.537
transition:emotionality	1, 89	0.01	0.87	.010	.353
propCorrect.z:transition:emotionality	1, 89	0.01	0.87	.010	.354
avoidance.z:transition:emotionality	1, 89	0.01	0.02	<.001	.892
attention:transition:emotionality	1, 89	0.01	0.03	<.001	.868
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.92	.010	.339
avoidance.z:attention:transition:emotionality	1, 89	0.01	3.42 +	.037	.068

For know

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.32	0.66	.007	.419
avoidance.z	1, 89	0.32	1.26	.014	.264
attention	1, 89	0.11	41.08 ***	.316	<.001
propCorrect.z:attention	1, 89	0.11	0.00	<.001	.991
avoidance.z:attention	1, 89	0.11	0.29	.003	.588
transition	1, 89	0.02	2.76 +	.030	.100
propCorrect.z:transition	1, 89	0.02	0.63	.007	.431
avoidance.z:transition	1, 89	0.02	0.90	.010	.346
emotionality	1, 89	0.01	0.18	.002	.668
propCorrect.z:emotionality	1, 89	0.01	0.44	.005	.507
avoidance.z:emotionality	1, 89	0.01	2.33	.026	.130
attention:transition	1, 89	0.02	2.51	.027	.116
propCorrect.z:attention:transition	1, 89	0.02	0.47	.005	.496
avoidance.z:attention:transition	1, 89	0.02	0.04	<.001	.846
attention:emotionality	1, 89	0.01	5.61 *	.059	.020
propCorrect.z:attention:emotionality	1, 89	0.01	0.87	.010	.353
avoidance.z:attention:emotionality	1, 89	0.01	0.07	<.001	.794
transition:emotionality	1, 89	0.01	0.50	.006	.482
propCorrect.z:transition:emotionality	1, 89	0.01	0.93	.010	.337
avoidance.z:transition:emotionality	1, 89	0.01	0.23	.003	.636
attention:transition:emotionality	1, 89	0.01	0.02	<.001	.899
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.22	.002	.642
avoidance.z:attention:transition:emotionality	1, 89	0.01	0.44	.005	.509

Avoidance was never sign.

Covariate Approach

For hits:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.17	0.42	.005	.520
approach.z	1, 89	0.17	0.14	.002	.707
attention	1, 89	0.05	499.36 ***	.849	<.001
propCorrect.z:attention	1, 89	0.05	4.32 *	.046	.041
approach.z:attention	1, 89	0.05	0.29	.003	.591
transition	1, 89	0.02	6.73 *	.070	.011
propCorrect.z:transition	1, 89	0.02	0.18	.002	.668
approach.z:transition	1, 89	0.02	0.31	.003	.580
emotionality	1, 89	0.01	2.26	.025	.136
propCorrect.z:emotionality	1, 89	0.01	0.93	.010	.338
approach.z:emotionality	1, 89	0.01	1.94	.021	.167
attention:transition	1, 89	0.02	16.92 ***	.160	<.001
propCorrect.z:attention:transition	1, 89	0.02	1.41	.016	.238
approach.z:attention:transition	1, 89	0.02	1.76	.019	.188
attention:emotionality	1, 89	0.02	11.86 ***	.118	<.001
propCorrect.z:attention:emotionality	1, 89	0.02	0.37	.004	.544
approach.z:attention:emotionality	1, 89	0.02	0.71	.008	.400
transition:emotionality	1, 89	0.02	0.13	.001	.720
propCorrect.z:transition:emotionality	1, 89	0.02	0.01	<.001	.941
approach.z:transition:emotionality	1, 89	0.02	0.10	.001	.754
attention:transition:emotionality	1, 89	0.01	0.00	<.001	.979
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.30	.003	.585
approach.z:attention:transition:emotionality	1, 89	0.01	0.53	.006	.468

For remember responses

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.25	0.09	<.001	.770
approach.z	1, 89	0.25	3.02 +	.033	.086
attention	1, 89	0.12	81.43 ***	.478	<.001
propCorrect.z:attention	1, 89	0.12	2.05	.023	.155
approach.z:attention	1, 89	0.12	2.45	.027	.121
transition	1, 89	0.01	0.95	.011	.331
propCorrect.z:transition	1, 89	0.01	1.98	.022	.163
approach.z:transition	1, 89	0.01	0.75	.008	.389
emotionality	1, 89	0.01	3.25 +	.035	.075
propCorrect.z:emotionality	1, 89	0.01	0.00	<.001	.976
approach.z:emotionality	1, 89	0.01	3.12 +	.034	.081
attention:transition	1, 89	0.01	10.36 **	.104	.002
propCorrect.z:attention:transition	1, 89	0.01	4.49 *	.048	.037
approach.z:attention:transition	1, 89	0.01	0.44	.005	.507
attention:emotionality	1, 89	0.01	2.22	.024	.139
propCorrect.z:attention:emotionality	1, 89	0.01	0.03	<.001	.872
approach.z:attention:emotionality	1, 89	0.01	0.34	.004	.560
transition:emotionality	1, 89	0.01	0.87	.010	.353
propCorrect.z:transition:emotionality	1, 89	0.01	0.79	.009	.378
approach.z:transition:emotionality	1, 89	0.01	0.21	.002	.647
attention:transition:emotionality	1, 89	0.01	0.03	<.001	.870
propCorrect.z:attention:transition:emotionality	1, 89	0.01	1.47	.016	.228
approach.z:attention:transition:emotionality	1, 89	0.01	0.40	.004	.528

For know:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.32	0.53	.006	.468
approach.z	1, 89	0.32	1.60	.018	.209
attention	1, 89	0.11	41.65 ***	.319	<.001
propCorrect.z:attention	1, 89	0.11	0.00	<.001	.991
approach.z:attention	1, 89	0.11	1.52	.017	.221
transition	1, 89	0.02	2.79 +	.030	.098
propCorrect.z:transition	1, 89	0.02	0.75	.008	.390
approach.z:transition	1, 89	0.02	1.80	.020	.183
emotionality	1, 89	0.01	0.18	.002	.672
propCorrect.z:emotionality	1, 89	0.01	0.92	.010	.341
approach.z:emotionality	1, 89	0.01	0.24	.003	.623
attention:transition	1, 89	0.02	2.53	.028	.115
propCorrect.z:attention:transition	1, 89	0.02	0.44	.005	.510
approach.z:attention:transition	1, 89	0.02	0.75	.008	.389
attention:emotionality	1, 89	0.01	5.61 *	.059	.020
propCorrect.z:attention:emotionality	1, 89	0.01	0.76	.009	.385
approach.z:attention:emotionality	1, 89	0.01	0.12	.001	.729
transition:emotionality	1, 89	0.01	0.50	.006	.483
propCorrect.z:transition:emotionality	1, 89	0.01	1.09	.012	.300
approach.z:transition:emotionality	1, 89	0.01	0.02	<.001	.879
attention:transition:emotionality	1, 89	0.01	0.02	<.001	.900
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.35	.004	.556
approach.z:attention:transition:emotionality	1, 89	0.01	0.01	<.001	.903

Covariate Self-Control

For hits:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.17	0.40	.004	.528
selfcontrol.z	1, 89	0.17	0.52	.006	.472
attention	1, 89	0.05	498.53 ***	.849	<.001
propCorrect.z:attention	1, 89	0.05	4.06 *	.044	.047
selfcontrol.z:attention	1, 89	0.05	0.14	.002	.706
transition	1, 89	0.02	6.73 *	.070	.011
propCorrect.z:transition	1, 89	0.02	0.12	.001	.726
selfcontrol.z:transition	1, 89	0.02	0.40	.004	.529
emotionality	1, 89	0.01	2.22	.024	.140
propCorrect.z:emotionality	1, 89	0.01	1.18	.013	.281
selfcontrol.z:emotionality	1, 89	0.01	0.22	.002	.642
attention:transition	1, 89	0.02	16.59 ***	.157	<.001
propCorrect.z:attention:transition	1, 89	0.02	1.19	.013	.279
selfcontrol.z:attention:transition	1, 89	0.02	0.02	<.001	.879
attention:emotionality	1, 89	0.02	11.88 ***	.118	<.001
propCorrect.z:attention:emotionality	1, 89	0.02	0.38	.004	.538
selfcontrol.z:attention:emotionality	1, 89	0.02	0.87	.010	.353
transition:emotionality	1, 89	0.02	0.13	.001	.717
propCorrect.z:transition:emotionality	1, 89	0.02	0.03	<.001	.860
selfcontrol.z:transition:emotionality	1, 89	0.02	1.76	.019	.188
attention:transition:emotionality	1, 89	0.01	0.00	<.001	.979
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.39	.004	.536
selfcontrol.z:attention:transition:emotionality	1, 89	0.01	0.11	.001	.743

For remember responses

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.26	0.03	<.001	.868
selfcontrol.z	1, 89	0.26	0.01	<.001	.943
attention	1, 89	0.12	80.14 ***	.474	<.001
propCorrect.z:attention	1, 89	0.12	1.56	.017	.215
selfcontrol.z:attention	1, 89	0.12	1.00	.011	.319
transition	1, 89	0.01	1.00	.011	.320
propCorrect.z:transition	1, 89	0.01	1.89	.021	.172
selfcontrol.z:transition	1, 89	0.01	5.12 *	.054	.026
emotionality	1, 89	0.01	3.15 +	.034	.079
propCorrect.z:emotionality	1, 89	0.01	0.02	<.001	.902
selfcontrol.z:emotionality	1, 89	0.01	0.17	.002	.680
attention:transition	1, 89	0.01	10.33 **	.104	.002
propCorrect.z:attention:transition	1, 89	0.01	4.17 *	.045	.044
selfcontrol.z:attention:transition	1, 89	0.01	0.20	.002	.653
attention:emotionality	1, 89	0.01	2.34	.026	.129
propCorrect.z:attention:emotionality	1, 89	0.01	0.06	<.001	.800
selfcontrol.z:attention:emotionality	1, 89	0.01	5.15 *	.055	.026
transition:emotionality	1, 89	0.01	0.91	.010	.341
propCorrect.z:transition:emotionality	1, 89	0.01	1.14	.013	.288
selfcontrol.z:transition:emotionality	1, 89	0.01	4.51 *	.048	.036
attention:transition:emotionality	1, 89	0.01	0.03	<.001	.867
propCorrect.z:attention:transition:emotionality	1, 89	0.01	2.03	.022	.157
selfcontrol.z:attention:transition:emotionality	1, 89	0.01	5.01 *	.053	.028

For know:

Effect	df	MSE	F	pes	p.value
propCorrect.z	1, 89	0.33	0.36	.004	.548
selfcontrol.z	1, 89	0.33	0.21	.002	.651
attention	1, 89	0.11	41.22 ***	.317	<.001
propCorrect.z:attention	1, 89	0.11	0.02	<.001	.901
selfcontrol.z:attention	1, 89	0.11	0.59	.007	.444
transition	1, 89	0.02	2.80 +	.030	.098
propCorrect.z:transition	1, 89	0.02	0.78	.009	.380
selfcontrol.z:transition	1, 89	0.02	1.96	.022	.165
emotionality	1, 89	0.01	0.18	.002	.672
propCorrect.z:emotionality	1, 89	0.01	0.85	.009	.360
selfcontrol.z:emotionality	1, 89	0.01	0.00	<.001	.986
attention:transition	1, 89	0.02	2.52	.028	.116
propCorrect.z:attention:transition	1, 89	0.02	0.48	.005	.492
selfcontrol.z:attention:transition	1, 89	0.02	0.36	.004	.551
attention:emotionality	1, 89	0.01	5.71 *	.060	.019
propCorrect.z:attention:emotionality	1, 89	0.01	0.97	.011	.327
selfcontrol.z:attention:emotionality	1, 89	0.01	1.71	.019	.195
transition:emotionality	1, 89	0.01	0.50	.006	.481
propCorrect.z:transition:emotionality	1, 89	0.01	1.23	.014	.270
selfcontrol.z:transition:emotionality	1, 89	0.01	0.72	.008	.397
attention:transition:emotionality	1, 89	0.01	0.02	<.001	.898
propCorrect.z:attention:transition:emotionality	1, 89	0.01	0.48	.005	.489
selfcontrol.z:attention:transition:emotionality	1, 89	0.01	3.00 +	.033	.087

For remember responses there were some sing. interactions with self control.

Check regression slopes:

## `geom_smooth()` using formula 'y ~ x'
## `geom_smooth()` using formula 'y ~ x'

Negative correlation between selfcontrol and memory selectivity?