Correlation between Trait Self-Control and Memory ratings

2022-06-05

Online Data

Method

Participants

173 participants completed the study (118 women and 55 men). Mean age was 19 years (SD = 1).

Material

Procedure

  1. In the study phase, participants completed a task switching procedure

  2. As a filler task, participants completed the letter-memory task.

  3. In the test phase, a surprise recognition memory test assessed participant’s memory for stimuli presented in the study phase

  4. In a final questionnaire Trait Self-Control was assessed.

Results

Hits and False Alarms

Overall the hit rate was 0.498 (SE = 0.014) and the false alarms rate was 0.206 (SE = 0.013).

Subtracting the false alarm rate from the hit rate gives us d’, which is the basis to compute the correlation with trait self-control.

Scatter plot with all participants:

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

## 
##  Shapiro-Wilk normality test
## 
## data:  df$selfcontrol
## W = 0.98835, p-value = 0.1646
## 
##  Shapiro-Wilk normality test
## 
## data:  df$d
## W = 0.93445, p-value = 4.315e-07

d’ is not from a normal distribution

Let’s look at the Q-Q plots

# selfcontrol
ggqqplot(df$selfcontrol, ylab = "selfcontrol")

# d
ggqqplot(df$d, ylab = "d'")

res <- cor.test(df$selfcontrol, df$d, 
                    method = "kendall")
res
## 
##  Kendall's rank correlation tau
## 
## data:  df$selfcontrol and df$d
## z = 0.36012, p-value = 0.7188
## alternative hypothesis: true tau is not equal to 0
## sample estimates:
##        tau 
## 0.01879384

check distribution of d’ (factor gender just for fun)

It would probably be wise to remove participants with very low d’ scores (<.1)

That’s better…

## Warning: geom_vline(): Ignoring `mapping` because `xintercept` was provided.
## Warning: geom_vline(): Ignoring `data` because `xintercept` was provided.

## 
##  Shapiro-Wilk normality test
## 
## data:  df$selfcontrol
## W = 0.98687, p-value = 0.201
## 
##  Shapiro-Wilk normality test
## 
## data:  df$d
## W = 0.99042, p-value = 0.4497

That’s fine.

This scatterplot for real (N = 141)

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

(This result corresponds to the output from Jamovi)

Hits for Targets

I believe the mean of target memory and distractor memory masks the relationship. Because, if you are good at selective processing then you would remember more targets than distractors (because attention is highly focused during encoding).

So let’s try this with only hits for targets. It’s also better because then we have the same number of trials used for computing hitrate and false alarm rate (there was a 2:1 ratio for the items old:new, as old items were divided into targets and distractors)

## 
##  Shapiro-Wilk normality test
## 
## data:  df$selfcontrol
## W = 0.98687, p-value = 0.201
## 
##  Shapiro-Wilk normality test
## 
## data:  df$d_target
## W = 0.97661, p-value = 0.01607

Ugh the d’ score for targets is again not normally distributed. Let’s check:

## Warning: geom_vline(): Ignoring `mapping` because `xintercept` was provided.
## Warning: geom_vline(): Ignoring `data` because `xintercept` was provided.

slightly skewed…

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

Z-standardization

Z-Standardization overall (d’)

Let’s check gender just for fun (with z-transformed-variables):

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

Follow-up analyses

Would it be of interest to look at switch vs. repeat trials separately? And also include the emotionality variable? Or separate by material (pictures vs. words)?

False alarm rate

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

Exploring Relationship with Remember - Know responses

This is also a kind of sensitivity measure…

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

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

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

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

Subscales of Trait Self-Control

Impulsivity

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

Restraint

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