Replication of Sleep Preferentially Enhances Memory for Emotional Components of Scenes by Payne, Stickgold, Swanberg and Kensinger (2008, Psychological Science)

Based on the prior write-up, describe any differences between the original and 1st replication in terms of methods, sample, sample size, and analysis. Note any potential problems such as exclusion rates, noisy data, or issues with analysis.

The first replication project differed from the original in that it recruited online crowd workers (MTurk), whereas the participants from the original study were college students at Harvard and Boston College. The replication project experiments were also run online, whereas the original experiment was done in person, in a lab.

In the original experiment, participants were randomly assigned to one of two possible conditions (one in which they would view a set of stimuli after sleeping and the other after 12 hours awake). In the replication, participants were allowed to self-select which condition they would be in. This presents a strong chance of selection bias, and also reduces our opportunity to make causal inferences from the results.

The replication attempted to recruit 48 participants, but only ended up being able to include 23 participants in the final analysis due to so many of them dropping out. The original study recruited 88 participants.

Methods

Power Analysis

How much power does your planned sample have for original effect? For an attenua

Planned sample size and/or termination rule, sampling frame, known demographics if any, preselection rules if any.

I plan to recruit at least 88 participants using prolific. This figure may change after power analyses are completed, however, seeing as this is a rescue attempt of a replication that may have failed largely due to a small sample size, I think it makes sense to plan for a sufficiently large sample. Participants who complete neither or only one of the two required experiment conditions will be removed from any analyses.

In the replication project, 91.3% of the sample had finished or attended college, so I will attempt to collect a sample that is more diverse in SES.

Materials

All materials - can quote directly from original article - just put the text in quotations and note that this was followed precisely. Or, quote directly and just point out exceptions to what was described in the original article.

“The scenes portrayed negative arousing or neutral objects placed on plausible neutral backgrounds. For each of 64 scenes (e.g., a car on a street), we created eight different versions by placing each of two similar neutral objects (e.g., two images of a car) and each of two related negative objects (e.g., two images of a car accident) on each of two plausible neutral backgrounds (e.g., two images of a street). An additional 32 scenes served as lures on a recognition memory test (Fig. 1). Participants in a previous study had rated the objects and backgrounds for valence and arousal, using 7-point scales (Kensinger, Garoff-Eaton, & Schacter, 2006). All negative objects had received arousal ratings of 5 to 7 (with high scores signifying an exciting or arousing image) and valence ratings lower than 3 (with low scores signifying a negative image). All neutral items (objects and backgrounds) had been rated as nonarousing (arousal values lower than 4) and neutral (valence ratings between 3 and 5).”

This will be followed precisely.

Fig. 1. “Examples of the scenes presented to subjects. Eight versions of each scene were created by combining each of four similar objects (two neutral objects, two negative and arousing emotional objects) with each of two plausible neutral backgrounds. In this example, the two neutral central objects are cars, and the two negative central objects are cars damaged in an accident; the neutral backgrounds are street scenes. Two of the eight versions of the completed scene are shown.” http://journals.sagepub.com/na101/home/literatum/publisher/sage/journals/content/pssa/2008/pssa_19_8/j.1467-9280.2008.02157.x/20160829/images/medium/10.1111_j.1467-9280.2008.02157.x-fig1.gif

Procedure

Can quote directly from original article - just put the text in quotations and note that this was followed precisely. Or, quote directly and just point out exceptions to what was described in the original article.

“Participants studied a set of 64 scenes (32 with a neutral object and 32 with a negative object, all on neutral backgrounds) for 5 s each, and then indicated on a 7-point scale whether they would approach or move away from the scene if they encountered it in real life. This task was used to maximize encoding.

After the delay period, participants performed an unexpected, self-paced recognition task. During this task, objects and backgrounds were presented separately and one at a time. Some of these objects and backgrounds were identical to the scene components that had been studied (e.g., the same car accident), others were the alternate version of the object or background and therefore shared the same verbal label but differed in specific visual details (e.g., a similar car accident), and others were objects or backgrounds that had not been studied (new). Participants never saw both the same and the similar version of an item at test. Each object or background was presented with a question (e.g., “Did you see a monkey?”). If the answer to the question was “yes,” participants pressed one button to indicate that the object or background was an exact match to a studied component (“same”) or a second button to indicate that it was not an exact match (“similar”). If the answer to the question was “no,” they pressed a third button.”

The recognition task includes 32 same objects (16 negative, 16 neutral), 32 similar objects (16 negative, 16 neutral), 32 new objects (16 negative, 16 neutral), 32 same backgrounds (16 previously shown with a negative object, 16 previously shown with a neutral object), 32 similar backgrounds (16 previously shown with a negative object, 16 previously shown with a neutral object), and 32 new backgrounds.”

This will be followed precisely. For the replication project, the experiment sessions before and after the delay period will be referred to as Session 1 and Session 2 respectively. I will also do this for my rescue attempt.

Controls

What attention checks, positive or negative controls, or other quality control measures are you adding so that a (positive or negative) result will be more interpretable?

Neither the original study, nor the replication controlled for sleep quality and number of hours slept, but I think it would be useful to control for this in the rescue attempt.

I also think it might be useful to periodically check with participants whether they actually find some of the stimuli to be negatively or neutrally arousing, which was the intention of the creators of the stimuli. After looking through the images that make up the stimulus set, whilst some of them may induce negative affect, some seemed too unrealistic to induce negative affect, or just seemed like they would stand out as being strange rather than negative (potentially due to appearing poorly photoshopped). This might need to be done with a separate set of participants in a different experiment.

Analysis Plan

Can also quote directly, though it is less often spelled out effectively for an analysis strategy section. The key is to report an analysis strategy that is as close to the original - data cleaning rules, data exclusion rules, covariates, etc. - as possible.

Clarify key analysis of interest here You can also pre-specify additional analyses you plan to do.

“We scored a response as specific recognition of visual details when a subject correctly responded”same” to a same item, but as general recognition without specific details when a subject responded “similar” to a same item. Because “similar” responses were constrained by the number of “same” responses (i.e., subjects responded “similar” only when they did not remember the visual details), we computed the general recognition score as the proportion of “similar” responses after exclusion of “same” responses (similar/[1- same]).” “Specific and general recognition scores were computed separately for central objects (negative or neutral) and for the peripheral neutral backgrounds (studied with either a negative or a neutral object).”

I will follow the replication, which planned to perform a 2 (condition: sleep, wake) x 2 (valence: negative, neutral) x 2 (scene component: object, background) mixed ANOVA. As well as a follow-up 2 (condition: sleep, wake) x 2 (valence: negative, neutral) mixed ANOVA applied on the recognition of objects and backgrounds separately.

Time and funds permitting, I also plan to do additional analyses double checking the valence ratings of the stimulus, as well as whether sleep quality (low, medium, high) has any potential interaction effect.

The term of interest is the three-way interaction (Condition x Valence x Scene Component). The original study found that negative, but not neutral stimuli were better remembered after sleep than wake. The original replaction did not find this result.

Differences from Original Study and 1st replication

Explicitly describe known differences in sample, setting, procedure, and analysis plan from original study. The goal, of course, is to minimize those differences, but differences will inevitably occur. Also, note whether such differences are anticipated to make a difference based on claims in the original article or subsequent published research on the conditions for obtaining the effect.

Key Differences

- I aim to recruit a sample greater than that of the original study (n = 88) and the first replication (n = 23) - I intend to recruit a sample diverse in SES - Participants will be randomly assigned to one of the two conditions (sleep or wake), rather than assigning themselves to a condition - The original study was done in-person, and replication was done online using Mturk workers. I will do my rescue online using prolific workers - I plan to run a smaller parallel study confirming the valence ratings of the stimuli by the original authors - I plan to ask participants in all conditions about their sleep quality the previous night when they complete the second of the two experimental sessions - I plan to include sleep quality in an additional mixed ANOVA. This may result in a slight change to the original claims of the article, which were that sleep preferentially enhances memory for emotional components of scenes. A novel result for this control would result in a new claim: Depending on sleep quality, sleep preferentially enhances memory for emotional components of scenes

Methods Addendum (Post Data Collection)

Actual Sample

I will be collecting data from 88 participants, which was the number of participants studied in the original paper.

Differences from pre-data collection methods plan

Any differences from what was described as the original plan, or “none”.

Results

Data preparation

Data preparation following the analysis plan.

####Import data #all data should be exported from qualtrics as "choice text", not numerical
#session 1 data 
data_s1 <- readSurvey("~/Downloads/PSYCH251 Session 1 - 2023_November 29, 2023_17.17.csv")

Warning: 'readSurvey' is deprecated.
Use 'read_survey' instead.
See help("Deprecated")

── Column specification ────────────────────────────────────────────────────────
cols(
  .default = col_double(),
  StartDate = col_datetime(format = ""),
  EndDate = col_datetime(format = ""),
  Status = col_character(),
  IPAddress = col_character(),
  Finished = col_logical(),
  RecordedDate = col_datetime(format = ""),
  ResponseId = col_character(),
  RecipientLastName = col_logical(),
  RecipientFirstName = col_logical(),
  RecipientEmail = col_logical(),
  ExternalReference = col_logical(),
  DistributionChannel = col_character(),
  UserLanguage = col_character(),
  ID = col_character(),
  `66_prac_rating` = col_character(),
  `67_prac_rating` = col_character(),
  `67_rating` = col_character(),
  `68_rating` = col_character(),
  `69_rating` = col_character(),
  `70_rating` = col_character()
  # ... with 62 more columns
)
ℹ Use `spec()` for the full column specifications.

#wake group data
data_s2_wake <- readSurvey("~/Downloads/PSYCH251 Session 2_wake - 2023_December 4, 2023_22.57.csv") %>% 
  mutate(condition = "wake")

Warning: 'readSurvey' is deprecated.
Use 'read_survey' instead.
See help("Deprecated")

── Column specification ────────────────────────────────────────────────────────
cols(
  .default = col_character(),
  StartDate = col_datetime(format = ""),
  EndDate = col_datetime(format = ""),
  Progress = col_double(),
  `Duration (in seconds)` = col_double(),
  Finished = col_logical(),
  RecordedDate = col_datetime(format = ""),
  RecipientLastName = col_logical(),
  RecipientFirstName = col_logical(),
  RecipientEmail = col_logical(),
  ExternalReference = col_logical(),
  LocationLatitude = col_double(),
  LocationLongitude = col_double(),
  code_s1 = col_double(),
  `Q42_First Click` = col_double(),
  `Q42_Last Click` = col_double(),
  `Q42_Page Submit` = col_double(),
  `Q42_Click Count` = col_double(),
  `Q43_First Click` = col_double(),
  `Q43_Last Click` = col_double(),
  `Q43_Page Submit` = col_double()
  # ... with 6 more columns
)
ℹ Use `spec()` for the full column specifications.

#sleep group data
data_s2_sleep <- readSurvey("~/Downloads/PSYCH251 Session 2_sleep 2023_December 4, 2023_23.09.csv") %>% 
  mutate(condition = "sleep")

Warning: 'readSurvey' is deprecated.
Use 'read_survey' instead.
See help("Deprecated")

── Column specification ────────────────────────────────────────────────────────
cols(
  .default = col_character(),
  StartDate = col_datetime(format = ""),
  EndDate = col_datetime(format = ""),
  Progress = col_double(),
  `Duration (in seconds)` = col_double(),
  Finished = col_logical(),
  RecordedDate = col_datetime(format = ""),
  RecipientLastName = col_logical(),
  RecipientFirstName = col_logical(),
  RecipientEmail = col_logical(),
  ExternalReference = col_logical(),
  LocationLatitude = col_double(),
  LocationLongitude = col_double(),
  code_s1 = col_double(),
  `Q42_First Click` = col_double(),
  `Q42_Last Click` = col_double(),
  `Q42_Page Submit` = col_double(),
  `Q42_Click Count` = col_double(),
  `Q43_First Click` = col_double(),
  `Q43_Last Click` = col_double(),
  `Q43_Page Submit` = col_double()
  # ... with 8 more columns
)
ℹ Use `spec()` for the full column specifications.

#combine the two raw datasets
data_s2 <- rbind(data_s2_wake,data_s2_sleep)

#### Data exclusion / filtering
# clean s1 data
data_s1_clean <- data_s1 %>% 
  select(-contains("Q34"), -contains("Q35"), -contains("Q19"), -contains("Q24"), -contains("prac")) #%>% 
  #filter(!is.na(am_or_pm))  

# final data collection started on 2018/12/12
#data_s1_clean <- data_s1_clean[mdy_hm(data_s1_clean$StartDate) >= ymd("2018-12-12"),]

# mark S1 variable names
names(data_s1_clean) <- str_c("s1_", names(data_s1_clean))

# clean s2 data
data_s2_clean <- data_s2 %>% 
  select(-contains("Q34"), -contains("Q35"), -contains("Q31"), -contains("Q42"), -contains("Q43"), -starts_with("Recipient"), -contains("s2p")) %>% 
  filter(!is.na("sleep_log#4_1"), 
         code_s1 != "999999", code_s1 != "99999", code_s1 != "888888")

# final data collection started on 2018/12/12
#data_s2_clean <- data_s2_clean[mdy_hm(data_s2_clean$StartDate) >= ymd("2018-12-12"),]

# categorize participants into sleep-condition and wake-condition
#data_s2_clean <- data_s2_clean %>% 
  #mutate(time = mdy_hm(StartDate), condition = ifelse(hour(time) >= 5 & hour(time) <= 12, "sleep", "wake")) # considering time difference, the sleep group (across the US) starts Session 2 between 5 am and 12 pm 

# join s1 and s2 data
data_s1s2 <- left_join(data_s2_clean, data_s1_clean, by = c("code" = "s1_code"))

# calculate the s1-s2 time gap
data_s1s2 = data_s1s2 %>% 
  mutate(int = interval(start = s1_StartDate, end =  StartDate)) %>% 
  mutate(time_gap_seconds = int_length(int)) %>% 
  mutate(time_gap_minutes = minute(seconds_to_period(time_gap_seconds))) %>%
  mutate(time_gap_hours = hour(seconds_to_period(time_gap_seconds))) %>% 
  select(-(int)) 

#data_s1s2 <- data_s1s2 %>% 
  #mutate(time_gap = as.numeric(mdy_hm(StartDate) - mdy_hm(s1_StartDate), units = "hours") ) # time gap in hours

# filter out participants who did not have S1 and S2 about 12 hours apart
#data_s1s2 <- data_s1s2 %>% 
  #filter(time_gap >10 & time_gap  < 14)
# check the condition assignment (is it consistent with reported am_or_pm during Session 1)


table(data_s1s2$condition, data_s1s2$s1_am_or_pm)

< table of extent 2 x 0 >

# distribution of S2 time of the day
ggplot(data_s1s2, aes(x = hour(StartDate), fill = condition))+
  geom_histogram(color = 'black') +
  labs(title = "Distribution of the time of Session 2 (in Pacific Time)", x = "Time of the day (hour)")+
  xlim(0, 24) + scale_y_continuous(breaks=seq(0, 10, 1))

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

Warning: Removed 4 rows containing missing values (`geom_bar()`).

# distribution of the time gap between S1 and S2
ggplot(data_s1s2, aes(x = time_gap_hours))+
  geom_histogram(color = 'black', fill = 'white') +
  labs(title = "Distribution of time gap between Session 1 and Session 2", x = "duration (hours)")+
  xlim(-1, 14)

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

Warning: Removed 2 rows containing non-finite values (`stat_bin()`).

#### Prepare data for analysis - create columns etc.
# temp1: data with Q1 gathered (Q1: did you see this object/background?)
data <- data_s1s2 %>% 
  gather(item, s2_q1, contains("s2_q1")) %>% 
  mutate(item = substr(item, 1, str_length(item)-3))

Warning: attributes are not identical across measure variables; they will be
dropped

# create item_number and rearrange the dataframe
data <- data %>% 
  mutate(item_number = as.numeric(str_sub(item, 1, str_length(item)-3))) %>%
  arrange(code_s1, item_number)


# read the stimuli list
item_list <- read_csv('~/Downloads/questions_list.csv')

Rows: 192 Columns: 6
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (5): item_filname, item_link, component, valence, type
dbl (1): item_number

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

# merge the item list to the data frame
data_joined <-  left_join(data, item_list, by = "item_number") 
# rearrange
data_joined <- data_joined %>% arrange(code_s1, item_number) 

## summarize into individual-wise data
data_ind <- data_joined %>% 
  # filter(type == "same") %>%  #only "same" trials were processed, one participant has 64 trials
  group_by(code_s1, condition, type, component, valence, s2_q1) %>% 
  mutate(count_ans = n()) #the number of all answers separately (same, similar, new)
data_ind <- data_ind %>% 
  # filter(type == "same") %>% 
  group_by(code_s1, condition, type, component, valence) %>% 
  mutate(count_all = n()) #the number of questions
# collapse to individual level
data_ind <- data_ind %>% 
  group_by(code_s1, condition, type, component, valence, s2_q1) %>% 
  summarise(count_ans = mean(count_ans, na.rm = T),
            count_all = mean(count_all, na.rm = T)) %>% 
  #filter(!is.na(s2_q1)) %>% #s2_q1 has to have a value
  spread(s2_q1, count_ans) #spread s2_q1

`summarise()` has grouped output by 'code_s1', 'condition', 'type',
'component', 'valence'. You can override using the `.groups` argument.

#replace na with 0
data_ind[is.na(data_ind)] <-  0 
#calculate specific and general recognition rate
data_ind <-  data_ind %>% 
  mutate(specific_recog = Identical/count_all, #specific recognition
         general_recog = Similar/(count_all-Identical) #general recognition
         ) 

# order the level of condition
data_ind$condition <- factor(data_ind$condition, levels=c('wake', 'sleep'), ordered=TRUE)
data_ind$component <- factor(data_ind$component, levels=c('object', 'background'), ordered=TRUE)

# show the head of the individual-wise data
kable(head(data_ind, 11), digit = 2)

code_s1	condition	type	component	valence	count_all	Identical	New	Similar	specific_recog	general_recog
287112	sleep	new	background	neutral	32	2	26	4	0.06	0.13
287112	sleep	new	object	negative	16	0	13	3	0.00	0.19
287112	sleep	new	object	neutral	16	0	12	4	0.00	0.25
287112	sleep	same	background	negative	16	1	12	3	0.06	0.20
287112	sleep	same	background	neutral	16	2	11	3	0.12	0.21
287112	sleep	same	object	negative	16	8	4	4	0.50	0.50
287112	sleep	same	object	neutral	16	3	9	4	0.19	0.31
287112	sleep	similar	background	negative	16	0	13	3	0.00	0.19
287112	sleep	similar	background	neutral	16	2	11	3	0.12	0.21
287112	sleep	similar	object	negative	16	0	14	2	0.00	0.12
287112	sleep	similar	object	neutral	16	3	9	4	0.19	0.31

###Confirmatory analysis

## 3-way anova: test on the 3-way interaction
anova_3way = ezANOVA(
    data = data_ind %>% filter(type == "same"),
    dv = .(general_recog),
    wid = .(code_s1), 
    within = .(component, valence),
    between = .(condition)
)

Warning: Converting "code_s1" to factor for ANOVA.

Warning: Converting "valence" to factor for ANOVA.

#Show the ANOVA & assumption tests.
print(anova_3way)

$ANOVA
                       Effect DFn DFd   F   p p<.05 ges
2                   condition   1   0 NaN NaN    NA   1
3                   component   1   0 NaN NaN    NA   1
5                     valence   1   0 NaN NaN    NA   1
4         condition:component   1   0 NaN NaN    NA   1
6           condition:valence   1   0 NaN NaN    NA   1
7           component:valence   1   0 NaN NaN    NA   1
8 condition:component:valence   1   0 NaN NaN    NA   1

## 2-way anova: test on the key 2-way interaction on object recognition
anova_2way_obj = ezANOVA(
    data = data_ind %>% filter(type == "same", component == "object"),
    dv = .(general_recog),
    wid = .(code_s1), 
    within = .(valence),
    between = .(condition)
)

Warning: Converting "code_s1" to factor for ANOVA.

Warning: Converting "valence" to factor for ANOVA.

print(anova_2way_obj)

$ANOVA
             Effect DFn DFd   F   p p<.05 ges
2         condition   1   0 NaN NaN    NA   1
3           valence   1   0 NaN NaN    NA   1
4 condition:valence   1   0 NaN NaN    NA   1

## 2-way anova: test on the 2-way interaction on background
anova_2way_bgd = ezANOVA(
    data = data_ind %>% filter(type == "same", component == "background"),
    dv = .(general_recog),
    wid = .(code_s1), 
    within = .(valence),
    between = .(condition)
)

Warning: Converting "code_s1" to factor for ANOVA.

Warning: Converting "valence" to factor for ANOVA.

print(anova_2way_bgd)

$ANOVA
             Effect DFn DFd   F   p p<.05 ges
2         condition   1   0 NaN NaN    NA   1
3           valence   1   0 NaN NaN    NA   1
4 condition:valence   1   0 NaN NaN    NA   1

###Data plots

# summarize data for plotting
data_summary <- data_ind %>% 
  filter(type == "same") %>% 
  group_by(condition, component,valence) %>% 
  summarise(mean_spe_recog = mean(specific_recog, na.rm = T), 
            sd_spe_recog = sd(specific_recog, na.rm = TRUE), 
            se_spe_recog = sd_spe_recog/sqrt(n()), 
            ci_spe_recog =  1.96*se_spe_recog,# qt(0.975,df=length(specific_recog)-1)
            mean_gen_recog = mean(general_recog, na.rm = T), 
            sd_gen_recog = sd(general_recog, na.rm = TRUE), 
            se_gen_recog = sd_gen_recog/sqrt(n()), 
            ci_gen_recog =  1.96*se_gen_recog, # qt(0.975,df=length(general_recog)-1)
            # mean_all_recog = mean(all_recog, na.rm = T), 
            # sd_all_recog = sd(all_recog, na.rm = TRUE), 
            # se_all_recog = sd_all_recog/sqrt(n()), 
            # ci_all_recog =  1.96*se_all_recog, # qt(0.975,df=length(general_recog)-1)
            count = n())

`summarise()` has grouped output by 'condition', 'component'. You can override
using the `.groups` argument.

# order the level of condition
# data_summary$condition <- factor(data_summary$condition, levels=c('wake', 'sleep'), ordered=TRUE)
# data_summary$component <- factor(data_summary$component, levels=c('object', 'background'), ordered=TRUE)

kable(head(data_summary, 8), digits = 2)

condition	component	valence	mean_spe_recog	sd_spe_recog	se_spe_recog	ci_spe_recog	mean_gen_recog	sd_gen_recog	se_gen_recog	ci_gen_recog	count
wake	object	negative	0.75	NA	NA	NA	0.50	NA	NA	NA	1
wake	object	neutral	0.25	NA	NA	NA	0.58	NA	NA	NA	1
wake	background	negative	0.19	NA	NA	NA	0.38	NA	NA	NA	1
wake	background	neutral	0.25	NA	NA	NA	0.17	NA	NA	NA	1
sleep	object	negative	0.50	NA	NA	NA	0.50	NA	NA	NA	1
sleep	object	neutral	0.19	NA	NA	NA	0.31	NA	NA	NA	1
sleep	background	negative	0.06	NA	NA	NA	0.20	NA	NA	NA	1
sleep	background	neutral	0.12	NA	NA	NA	0.21	NA	NA	NA	1

The results of general recognition of the two groups (wake and sleep) is in the subplot A blow.

# plot general recognition
p1 <- ggplot(data_summary , aes(x = condition, y = mean_gen_recog, fill = component))+
  geom_bar(color = "black", stat = "identity", position=position_dodge(), width=0.5) + 
  geom_errorbar(aes(ymin=mean_gen_recog - ci_gen_recog, ymax=mean_gen_recog + ci_gen_recog), position=position_dodge(.5), width=.2, size=1) +
  #geom_point(data = data_ind, aes(x = condition, y = general_recog, fill = component), position=position_dodge(.5), color = "darkred", alpha = .5)+
  facet_grid(. ~ valence)+
  labs(title = "General recognition")+
  scale_fill_grey()+ ggthemes::theme_few()

Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
ℹ Please use `linewidth` instead.

# plot specific recognition
p2 <- ggplot(data_summary , aes(x = condition, y = mean_spe_recog, fill = component))+
  geom_bar(color = "black", stat = "identity", position=position_dodge(), width=0.5) + 
  geom_errorbar(aes(ymin=mean_spe_recog - ci_spe_recog, ymax=mean_spe_recog + ci_spe_recog), position=position_dodge(.5), width=.2, size=1) +
  #geom_point(data = data_ind, aes(x = condition, y = general_recog, fill = component), position=position_dodge(.5), color = "darkred", alpha = .5)+
  facet_grid(. ~ valence)+
  labs(title = "Specific recognition")+
  scale_fill_grey()+ ggthemes::theme_few()

plot_grid(p1, p2, ncol = 1, labels = c('A', 'B'))

Results of control measures

How did people perform on any quality control checks or positive and negative controls?

Exploratory analyses

Any follow-up analyses desired (not required).

Discussion

Mini meta analysis

For the 3-way interaction between condition, object and valence - the original study had an F statistic of 13.1 and a p-value of 0.001 - the original replication had an F statistic of 0.8344312 and p-value of 0.371363404.

Combining across the original paper, 1st replication, and 2nd replication, what is the aggregate effect size?

Summary of Replication Attempt

Open the discussion section with a paragraph summarizing the primary result from the confirmatory analysis and the assessment of whether it replicated, partially replicated, or failed to replicate the original result.

Commentary

Add open-ended commentary (if any) reflecting (a) insights from follow-up exploratory analysis, (b) assessment of the meaning of the replication (or not) - e.g., for a failure to replicate, are the differences between original and present study ones that definitely, plausibly, or are unlikely to have been moderators of the result, and (c) discussion of any objections or challenges raised by the current and original authors about the replication attempt. None of these need to be long.