ideas and concepts

Introduction to the class

• Install R in your class computer
  • Refer to this page

• Week 01 materials

• Theories 참조
• 연구문제와 가설 참조
• 연구문제 (research question)
• 개념 (concept) vs 개념 (concept)
  • concepts, ideas, etc.

Assignment

ideas and concepts

• HOW?

• ROSENBERG SELF-ESTEEM SCALE

연구문제와 가설 참조
위의 문서를 꼭 읽어야 합니다.
연구설계

• 연역 대. 귀납 (induction vs. deduction)
• 연구문제 (research question)
  • Conceptualization
  • Operationalization
• 가설 (hypothesis)
  • 차이와 연관 가설 참조
  • Difference
  • Association
• 변인 (variable) 설명
  • 속성 (attributes): 측정수준: 척도의 4가지 유형 참조 (p.78).
  • 종류와 숫자 (교재 참조: 불연속 vs. 연속)
  • NOIR (교재 참조: 명명, 서열, 간격, 비율척도)
• Types of Variable 변인의 종류
  • Dependent
  • Independent
  • Control
  • Moderating (Intervening)
  • eg.,
    • IV, DV: 부모의 교육수준 –> 자녀의 수능점수
    • moderating: 부모의 수입수준의 개입
    • control: 부모의 교육수준 (대학원 이상으로 콘트롤, 15년 이상)
    • intervening: 학생의 성별
• How to use data to make a hit TV show

• 그렇다면 어떤 가설과 연구문제가 적절한가?
• Sampling
  • Sampling
  • sampling frame
  • ECOBS
  • Probability sampling
    • Systematic Sampling:
    • Stratified Sampling:
    • Multistage Cluster Sampling:
    • Stratified in Multistage Cluster Sampling:
  • Non-probability sampling
  • Sample frame
  • Sample vs Population

Assignment 2

1. Please write hypotheses like the ones that we discussed in the class.
2. And describe how you are going to collect data and how you will be summarizing the outcome.

ideas and concepts

For the lecture content

Assignment

ideas and concepts

Ch. 5, 6, 7, 8

• range 혹은 범위
• interquartile range 혹은 사분위 범위
• 평균편차
• 변량 Variance 혹은 분산
  • 모집단분산
  • 표본분산 why n-1 degrees of freedom 혹은 df
  • text 부분: 추정치로서의 평균과 변량(분산)
• 표준편차 Standard Deviation
• 계산공식 분산계산공식

optical_illusion <- c(1.73, 1.06, 2.03, 
1.40, .95, 1.13, 1.41, 1.73, 1.63, 1.56)

> mean(optical_illusion)
[1] 1.463

> var(optical_illusion)
[1] 0.1160678

> sqrt(var(optical_illusion))
[1] 0.3406872

> stem(optical_illusion)

  The decimal point is at the |

  0 | 9
  1 | 1144
  1 | 6677
  2 | 0

> median(optical_illusion)
[1] 1.485

> mode(optical_illusion)
[1] "numeric"

• 정상분포 (정규분포) Normal Distribution
• 표준점수 z score

finger_tap<-rnorm(n=10000, m=59, sd=7) 
hist(finger_tap)

see normal distribution table
or use R (qnorm, pnorm)
.05 (5%)에 들때의 z score? = -1.645

qnorm(0.95)
[1] 1.644854

(혹은 .025(2.5%)일 경우 = -1.96 =about -2)

qnorm(0.975)
[1] 1.959964

그렇다면, 이 환자의 z score 는?
z-score = (45-59) / 7 = -2

z=-2 일때 분포확률 = 2.28% = 0.0228

> pnorm(-2)
[1] 0.02275013

How to look at “Normal” person? within .05? or .01? = 기각수준, 유의도수준

Please note that this is a hypothetical test for one individual (not a sample) against population.

• 가설검증 (Hypothesis testing)
  • 영가설(null hypothesis)
  • 연구가설, 대립가설 (research hypothesis, alternative hypothesis)
• types of error
  • type I error
  • type II error

• central limit theorem
  • sampling distribution
  • standard error

• z-test
• t-test

YSR (Youth Self-Report Inventory) 의 우울/불안 척도 ( $ \mu = 50, \sigma = 10 $ )
다섯 아이의 평균 = 56 일때, 이 아이들의 정상정도에 대한 가설 검증

• 절차를 옆사람에게 설명하시오.

> 10/sqrt(5)
[1] 4.47

Assignment

1. The more time people spend using social media, the less they read books.
2. Drinking energy drinks makes people more aggressive.
3. Taking a nap in the afternoon makes people more focused for the rest of the day.
4. Spending time with a family dog decreases the amount of stress someone is feeling.
5. Eating breakfast in the morning increases the ability to learn in school.

ideas and concepts

Week05 materials

Assignment

Quiz 1 on the next Wednesday

ideas and concepts

Quiz on April 11
ANOVA
Repeated Measure ANOVA
Factorial ANOVA

Assignment

시험범위

Mid term 범위:

• Ch. 1, 2, 3, 4, 5, 6, 7, 8
• + z-test

Mid term 이후 시험:

• Ch. 5, 6, 7, 8 +
• Ch. 12, 13, 14, 15 (효과크기 15.3 이후 제외), 16

ideas and concepts

Lecture content

Assignment

Mid-term period

ideas and concepts

Factorial ANOVA
Repeated Measure ANOVA – in a future week

Assignment

ideas and concepts

Children's day
Budah Birthday

Assignment

ideas and concepts

Correlation
Regression

Assignment

ideas and concepts

correlation
Regression

• Variance = SS_total / df
• SS_tot = sum of error squared predicted by mean alone
• SS_residual
  • Regression line
  • a and b in $ \hat{Y} = a + b X $
    • $b = {SP} / {SS_{X}}$
    • $a = \overline{Y} - b {\overline{X}}$
  • error squared predicted by regression line
• SS_regression = error squared overcome by regression line
• SS_tot = SS_regression + SS_residual
• If SS_regression is big enough, we can say
  • X's contribution to explain y's variation is significant
  • How to determine that? → F test
• $\text{F test} = MS_{\text{regression}} / MS_{\text{total}} $
  • with $\text{df}_{\text{regression}} = k - 1$ ; and
  • $\text{df}_{\text{total}} = n - 1$
• $\text{R}^{\text{2}} = \text{SS}_{\text{reg}} / \text{SS}_{\text{tot}}$
• will be clear with multiple regression
  • degrees of freedom을 고려한 R² = adjusted R²
    • addition of IVs will always increase R².
    • should be penalized (or adjusted)
    • so, when R² = 1 - (SS_res/SS_tot), use
      • SS_res → SS_res/df_res
      • df_res = n - p - 1
      • p = number of IVs
      • if p increases, the calculated value will be decreased, which will give you adjusted R² value.
      • SS_tot → SS_total/df_tot
      • df_tot = n - 1
  • meaning of t test for slope b
    • Suppose that in $ \hat{Y} = a + b_{1} X_{1} + b_{2} X_{2} $, Xs are not correlated to each other, and X is not contributing anything to Y's variance,
    • we can say that b = 0.
    • This is a null hypothesis for testing b
    • Actual test for determining the contribution of bs is t-test
      • t = b1 - b / SE_b
      • $\displaystyle \text{SE}_{\text{b}} = \frac{s_{\text{est}}}{\sqrt{SSX}}$

Multiple Regression
Sequential Regression
Using Dummy Variables

t-test in r
anova

• oneway anova in r
• twoway anova in r

linear regression in r
multiple regression in r

Assignment

ideas and concepts

Assignment

Announcement

Quiz: 6/4

• t-test
• F-test
  • ANOVA
  • Factorial ANOVA
  • Repeated Measures ANOVA
  • Effect size for ANOVA
• Regression
  • correlation
  • regression
  • multiple regression

기본적으로 위를 포함하지만, 위를 이해하기 위해서는

• standard deviation
• variance
• central limit theorem
  • sampling distribution
  • standard error
• hypothesis testing
• z-test
• types of error
• variable
• types of variable 등등을 이해해야 합니다.

또한 위를 포함하는 교재의 범위는

• Ch 4: one sample t-test
• Ch 5: paired sample t-test
• Ch 7: two sample t-test
• Ch 12: one-way ANOVA
• Ch 14: two-way ANOVA
• Ch 9: correlation analysis
• Ch 10: simple linear regression
• Ch 11: multiple regression

6일 현충일
In continuation with ANOVA, Factorial ANOVA
Repeated Measures ANOVA
post hoc test
Effect size for ANOVA

13일, 지방선거일
week15

Quiz: 지난 번 범위 + regression 부분 일체

• t-test
• F-test
  • ANOVA
  • Factorial ANOVA
  • Repeated Measures ANOVA
  • Effect size for ANOVA
• regression
  • regression
  • multiple regression
    • 무엇부터_라는_문제와 determining_ivs_role 부분 포함.
  • using dummy variables: 기본적인 논리를 중심으로 이해하세요.

기본적으로 위를 포함하지만, 위를 이해하기 위해서는

• standard deviation
• variance
• central limit theorem
  • sampling distribution
  • standard error
• hypothesis testing
• z-test
• types of error
• variable
• types of variable 등등을 이해해야 합니다.

또한 위를 포함하는 교재의 범위는

• Ch 12: 신뢰한계에 대해서는 수업중에 다루지 않았으므로 마지막 퀴즈에서도 제외합니다.
• Ch 13, Ch 14
• Ch 15:
• Ch 16:
  • 단일하지 않은 표본크기 포함
  • 다중비교(post hoc 혹은 multiple comparison techniques) 포함 (단 퀴즈에서 수학적인 것은 다루지 않습니다).
  • 효과크기 중 에타제곱에 해당하는 부분만 포함 (에타제곱, 파샬에타제곱, 오메가)
  • 결과보고하기 포함 (다루지 않았으나 숙지하시기 바랍니다)
• Ch 17 (factorial)
  • 효과크기 중 에타제곱이 아닌 부분은 (r-가족, 오메가 제곱 등) 제외
  • 17.7 제외
  • 17.8, 17.9 포함
• Ch 18 (repeated measures anova)

Final-term