Beruflich Dokumente
Kultur Dokumente
TUGASAN KUMPULAN :
KUMPULAN RAUDATUL JANNAH
AHLI KUMPULAN :
1.
- KORELASI
2.
- T-TEST BERKEMBAR
3.
- ANOVA
4.
5.
- T-TEST
METODS
ANALISIS KORELASI
ISSUES
1) What is the level of stress, self-esteem and substance abuse?
2) What is the relationship between stress, self-esteem and substance
abuse?
3) Does self-esteem mediate the relationship between stress and
substance abuse?
OBJECTIVE
The present study assessed the mediating effect of self-esteem in the
relationship between stress and substance abuse among adolescents.
HIPOTESIS
There is a relationship between self-esteem effects of stress and drugs
abuse among adolescents.
RESULT
INTERPRESTASI
---------------------------------------------------------------------------------------Pretest
Subscale
Attitude
4.15
Knowledge
Skills
SD
Posttest
M
SD
.26
4.53
.26
2.34
.77
4.02
.52
2.38
.68
3.99
.41
TYPE OF TEST
This
Comparisons
Tukeys
RESULT ANOVA
Table 1
N
All Subjects
Face-to-face
Blended
Online
167
58
46
63
Number
Correct
25.3
28.9
22.0
24.3
SD for Number
Correct
14.72
10.83
16.61
15.83
Percentage
52.6%
60.2%
45.8%
50.7%
SD for
Percentage
30.69
22.61
34.59
33.02
INTERPRETASI ANOVA
Note. The N column lists the number of subjects in the group. The
number correct is out of 48 problems. For Fall 2009, the success
rate across all 104 Intermediate Algebra courses at this institution
was 54.7%. The standard deviation (SD) for both the number correct
and percentage is provided.
For illustrative purposes, effect sizes were calculated for the Course
Average comparisons. The effect size for the face-to-face vs. online
difference was 0.17 favoring the face-to-face environment. For the
face-to-face vs. blended comparison, the effect size was 0.53, also
favoring the face-to-face environment. The online vs. blended
comparison effect size was 0.31, favoring the online environment.
Table
2
Mean and Standard Deviation Percent Correct on Unit Tests, IACE, and Course Average
by Learning Environment (Complete Dataset N = 167)
Face-toface
(N = 58)
Blended (N = 46)
Online (N = 63)
F(2, 164)
Pairwise tests of
significanceh
70.1 (21.4)
67.9 (32.1)
77.6 (24.6)
2.21
0.113
88.7 (14.7)
69.5 (34.5)
75.0 (32.5)
6.54
0.002*
50.4 (23.1)
45.4 (32.2)
59.3 (30.3)
3.39
0.036*
67.3 (22.7)
55.6 (33.5)
64.2 (34.1)
2.00
0.138
83.4 (20.2)
62.2 (41.1)
78.2 (35.4)
5.68
0.004*
63.8 (26.9)
52.8 (38.4)
64.7 (37.7)
1.82
0.165
75.5 (27.0)
54.4 (43.3)
65.0 (43.5)
3.86
0.023*
F>B
60.2 (22.6)
68.1 (18.7)
45.8 (34.6)
54.5 (32.0)
50.7 (33.0)
63.9 (28.5)
3.13
3.40
0.046*
0.036*
F>B
F>B
Unit Test
#1a
F>B;F>O
Unit Test
#3c
Unit Test
#5e
Unit Test
#7g
IACE
Course Average
Table 3
Mean and Standard Deviation Percent Correct on Unit Tests, IACE, and Course
Average by Learning Environment (Attrition Adjusted Dataset N=134)
Face-toface
(N = 54)
Blended (N = 32)
Online (N = 48)
F(2, 131)
72.2 (19.5)
79.3 (21.1)
85.7 (11.4)
7.53
0.001*
90.7 (8.8)
86.6 (12.8)
88.4 (10.4)
1.63
0.200
52.7 (21.5)
59.8 (26.1)
70.4 (21.5)
7.71
0.001*
70.4 (19.0)
72.2 (20.7)
78.2 (20.1)
2.07
0.130
87.5 (13.3)
83.9 (20.4)
92.4 (15.3)
2.84
0.062
67.2 (24.1)
73.9 (22.6)
81.6 (21.7)
5.04
0.008*
Pairwise tests of
significanceh
O>F
Unit Test
#1a
Unit Test
#3c
Unit Test
#4d
Unit Test
#6f
O>F
78.2
(28.3)
81.4
(31.9)
0.14
0.869
IACE
64.7
(16.0)
65.8
(19.4)
66.5
(19.1)
0.13
0.878
Course
71.6
(13.3)
73.1
(15.8)
78.1
(10.1)
One discrepancy among the age and gender controlled analyses was
that the Online vs. Face-to-face difference was no longer statistically
significant after control of multiple comparisons by Tukeys HSD
(difference = 5.3 SE = 2.8, t = 1.93; adjusted p-value = 0.135).
METOD
Jumlah
jenis instrumen
i)
ii)
Skala
Likert
Jadual 2:
Perhubungan antara Tekanan Pekerjaan
dan Big Five Trait Personaliti
Extraversion
Keterbukaan
Bersetuju
Sifat berhati-hati
Stress kerja
Keputusan ini menunjukkan bahawa jumlah tekanan kerja yang signifikan dengan extraversion
(r=0.849, kesan yang kuat), keterbukaan (r=0.886, kesan yang kuat), agreeableness (r=0.859,
Kesan yang kuat), sifat berhati-hati (r=0.639, kesan sederhana) dan dengan stres kerja
(r=0.401, kesan sederhana).
REGRESSION:
OBJEKTIF :
Menentukan faktor personaliti yang
menjadi peramal kepada tekanan kerja.
PERSOALAN :
Adakah terdapat pengaruh personaliti
terhadap tekanan kerja?
HIPOTESIS :
Sifat-sifat personaliti yang menjadi
faktor peramal kepada tekanan kerja
Tekanan
Kerja
Personaliti
Jadual 3 :
Profil Demografi
Tekanan
Kerja
Personaliti
Jadual 4 :
Khi Kuasa Dua antara Profil Demografi
dan Tekanan Pekerjaan
Jadual 5 :
Khi Kuasa Dua antara Profil Demografi
dan Personaliti
ANALISIS T-TEST
Using the Students t-test with
extremely small sample sizes
(Using the Students t-test with
extremely small sample sizes Volume
18, Number 10, August 2013)
Disediakan oleh :
FATEN NAJUA BT. AHMAD
METHOD
One of the two distributions was shifted with a value of Dwith respect to 0.
A sample was drawn from the population, and submitted to either the onesample t-test with a reference value of zero, or to the two-sample t-test
for testing the null hypothesis that the populations have equal means.
The simulations were carried out for Ds between 0 (i.e., the null
hypothesis holds) and 40 (i.e., the alternative hypothesis holds with an
extremely large effect), and for N = 2, N = 3, and N = 5. In the twosample
t-test, both samples were of equal size (i.e., N = M). A p value below 0.05
was considered statistically significant. All analyses were two-tailed. Each
case was repeated 100,000 times.
OBJECTIVE
1. behavior of the Welch test and a rank-transformation prior to conducting the t-test (t-testR).
2. Simulations were conducted to determine the statistical power and Type I error rate of the one
sample and two-sample t-tests.
3. Sampling was done from a normally distributed population with a mean of 0 and a standard
deviation of 1. One of the two distributions was shifted with a value of D with respect to 0. A sample
was drawn from the population, and submitted to either the one-sample t-test.
ISSUES
1.The present simulation study estimated the type 1 error rate and statistical power of the one- and
two sample t tests for normally distributed populations and various distortions such as unequal
sample sizes , unequal variences, the combination of unequel sample sizes and unequel variences
and lognormal population distribution
HIPOTESIS
There is difference type one error rate and statistical power of the one- and two sample t tests for
normally distributed populations.
RESULTS
DISCUSSION
The present simulation study showed that there is no fundamental objection to using a
regular t-test with extremely small sample sizes. Even a sample size as small as 2 did not
pose problems. In most of the simulated cases, the Type I error rate did not exceed the
nominal value of 5%. A paired t-test is also feasible with extremely small sample sizes,
particularly when the within-pair correlation coefficient is high.
A high Type I error rate was observed for unequal variances combined with unequal sample
sizes (withthe smaller sample drawn from the high variance population), and for a onesample t-test applied to nonnormally distributed data. The simulations further clarified that
when the sample size is extremely small,
Type II errors can only be avoided if the effect size is extremely large. In other words,
conducting a t-test with extremely small samples is feasible, as long as the true effect size is
large
The fact that the t-test functions properly for extremely small sample sizes may come as no
surprise to the informed reader. In fact, William Sealy Gosset (working under the pen
nameStudent) developed the t-test especially for small sample sizes (Student, 1908; for
reviews see Box, 1987; Lehmann, 2012; Welch, 1958; Zabell, 2008), a condition where the
traditional ztest provides a high rate of false positives. Student himself verified his tdistribution on anthropometric data of 3,000 criminals, which he randomly divided into 750
samples each having a sample size of 4.