• What are the colors on French flag?
• What are the colors on Belgiens flag?

• What is the first idea show up in your mind?

1. Preview the original study
2. Post-review the critical results
3. Actions in this project

• Public issue (google "Google Effect")
• Expand the scope of cognitive psychology
• Check the original article

• "Once information has been accessed, our internal encoding is increased for where the information is to be found rather than for the information itself."
• Exp 1: If we are asked a hardly answered question, do we immediately think to go online to find out?
• Exp 2: Do we remember information we expect to have later access to — as we could search online?
• Exp 3: Do we remember where to find information that is available online?
• Exp 4: Do we remember where to find than what is the information?

• Exp 1: Modification Stroop Task– after answered easy/hard questions
• Exp 2: Recall task– Saved/Erased studied notes; cue/no cue to remeber the typings
• Exp 3: Recognition task– Saved/Categorized/Erased studied notes
• Exp 4: Recall task– Firstly recall the studied notes; secondly recall the folders saved the notes.

1. Computer terms had a Stroop effect between easy and hard questions
2. Erased notes(30%) had a higher accuracy than saved notes(20%)
3. Erased notes (93%) > Saved(88%) = Categorized(85%)
4. Where to saved (49%) had a higher recall accuracy than what to saved(23%)

• External factor is aid or balk to human memory?
• Is the knowledge representation matter?
• Have we to rethink how human memory operate?

Will we replicate the results of experiment 1?

• Understand within-participant design
  
• Learn how to evaluate the sample size
  
• How to make exact replication

• Randomly assign the Questions and critical words
• Take a look at "GoogleEffect_V1"

• Stimulus: Question lists; Critical words;
• Modified Stroop Task (we will experience later)

• Structure of questions and critical words
• Translation
• Match cultural situation

• Procedure of the modified Stroop task
• codebook of the raw data
• Test "GoogleEffect_V1"

1. Estimate the effect size
2. Compute the power by effect size
3. Decide the sample size by the effect size

esComp() in furncions.r
We have this function from [RP:P project]

esComp <- function(
  x,
  df1,
  df2,
  N,
  esType){
  esComp <- ifelse(esType=="t",           # For t test
                   sqrt((x^2*(1 / df2)) / (((x^2*1) / df2) + 1)),
                   ifelse(
                     esType=="F",         # For F test
                     sqrt((x*(df1 / df2)) / (((x*df1) / df2) + 1))*sqrt(1/df1),
                     ifelse(
                       esType=="r",       # For correlation
                       x,
                       ifelse(
                         esType=="Chi2",  # For chi square
                         sqrt(x/N),
                         ifelse(
                           esType == "z", # For z test
                           tanh(x * sqrt(1/(N-3))), 
                           NA
                         )
                       )
                     )
                   ))
  return(esComp)
}

1. Computer terms vs. General terms: t(68) = 3.26, N = 69
2. Computer terms after easy vs. after hard: t(68) = 2.98, N = 69

## [1] 0.3676462

## [1] 0.3398666

1. Computer terms vs. General terms
2. Computer terms after easy vs. after hard
3. What is the major effect?

## [1] 0.8532638

## [1] 0.7947758

How large is the major effect?

• Could these treatments violate research ethic?
• Without harm for the participants after the study
• Considerations for IRB

• Nearly exact replication of Experiment 1
• Except the languages, all the details in the method are as the same as the original study.
• Considering the power, how to decide the acceptable sample size?
• Modified Stroop Task came from Emotional Stroop Test

• Lower the difference of languages and cultures
• Categories of questions
• Equivalent terms

• Introduction: Autopsy of the original study

1. Hypothesis of Experiment 1
2. Design: Independent variable and Dependant variable
3. Prediction

• Method

1. Sample: how many participant?
2. Treatments: Questions(Easy, Hard), Critical Terms(Computer, General) <- How to translate/modify?
3. Measurement: Reaction times
4. Proceduree: Modified Stroop Task
5. Analysis: coodbook <- check the csv file (after complete GoogleEffectV1)