PENELITIAN

HYPOTHETICO DEDUCTIVE
METHOD
Prof. Dr. Imam Ghozali, M.Com
Program S3 Ilmu Ekonomi
Universitas Diponegoro
 BERPIKIR INDUKTIF DAN DEDUKTIF

“ Metal Dipanaskan memuai”

Universal Statement
TEORI

Induktif Deduktif
Normatif & Positif

Observasi Empiris Describe (deskripsi)

Singular Statement Explain (menjelaskan)
Predict (meramalkan)
 HUBUNGAN TEORI DAN PENELITIAN

Sifat Teori
Jenis Penelitian Deskriptif Explain Prediksi
1. Exploratory Non-Empiris
Filsafat, Sejarah
Empiris
Study Kasus,survey
Etnography

2. Uji Hipotesis Empiris

a. Korelasi Survey
Non-eksperimen
b. Regresi Empiris
Eksperimen
Survey
 THE STEPS IN HYPOTHETICO DEDUCTIVE METHOD

Observation
4
3 5 6 7
Theoretical Research Data
Problem Hypotheses
Framework Design Collection
Definition
Analysis

Preliminary
Data Deduction
Gathering
2 8
 Problem Definition
1. Latar Belakang Masalah (7- 10 hal)
- Theory Gap atau Research Gap
- Masalah Penelitian dapat berasal dari
a. Pengamatan empiris
b. Penelitian sebelumnya: replikasi,
modifikasi atau pengembangan
 Problem Definition
2. Perumusan Masalah
- Diawali dengan problem statement (pernyataan
masalah)
- Diwujudkan dalam research question (pertanyaan
penelitian) : What, How dan Why
3. Tujuan Penelitian
- Menjawab masalah penelitian
- Gunakan kata kerja dgn kognitif tinggi (menganalisis,
mengevaluasi)
- Harus konsisten antara judul penelitian, latar
belakang masalah, perumusan masalah, tujhuan
penelitian dan hipotesis
A Comparison of Accounting Standards Setting in The International Environment
And in The United States

Purpose of The Study

The primary purpose of this research study is to investigate the extent to which
Experience in establishing accounting standards in the United States can be
Utilized in the developing the process of establishing international accounting
Standards for the financial statement reporting of enterprises operating in the
International environment.
Research Questions

The research was undertaken to answer the following primary research

Questions:
What need for modifications to the process of establishing
accounting standards in the international environment can
be inferred from a comparison with the process of establishing
accounting standards in the united states?
The subsidiary research questions are:
1. How are accounting standards currently established for the
financial statement reporting of profit oriented United States
enterprises?

2. How are international accounting standards currently established

for the financial statement reporting of profit oriented enterprises?

3. What are the similarities and differences in the process of

establishing accounting standards in the United States and on
an international level?

4. What differences exist in the environment within which

financial statement reporting takes place on an international
level when compared with the United States?
 Some common mistakes in problem
formulation
1. Collecting data without a well defined plan or purpose,
hoping to make some sense out of it afterward.
2. Taking a “batch of data” that already exist and
attempting to fit meaningful research questions to it.
3. Defining objectives in such general or ambiguous term
that your interpretations and conclusions will be
arbitrary and invalid.
4. Undertaking a research project without reviewing
professional literature on the subject.
5. Ad hoc research, unique to a given situation, permitting
no generalizations beyond the situation itself and
making no contribution to the general body of
educational research
 Some common mistakes in problem
formulation
6. Failure to base research on a sound theoretical or conceptual
framework, which would tie together the divergent masses of
research into a systematic and comparative scheme, providing
feed back and evaluation of educational theory.
7. Failure to make explicit and clear the underlying assumptions within
your research so that it can be evaluated in terms of these
foundations.
8. Failure to recognized the limitations in your approach, implied or
explicit, that place restrictions on the conclusion and how they
apply to other situations.
9. Failure to anticipate alternative rival hypotheses that would also
account for given set findings and which challenge the
interpretations and conclusions reached by the investigator.
 KERANGKA PEMIKIRAN TEORITIS
Suatu conceptual model untk menggambarkan hubungan antar variabel yg
Telah diidentifkasi sebagai faktor penting yg berkaitan dg masalah penelitia

Content of Theory

Concepts Definiton Proposition

Abstract idea State meaning of the Statement about concept
concept Or relationship between
concept

Non-Variable Variable

Observe Un-observed

Kinerja (ROA)
Inflasi (IHK)
 Skala Pengukuran Variabel
Skala Kategori Peringkat Jarak Unique Nilai Dispersi Test
Origin Tengah signif

Nominal Ya Tdk Tdk Tdk Mode X2

Ordinal Ya Ya Tdk Tdk Median Inter- Rank

quartil Order
Deviasi Corr
Interval Ya Ya Ya Tdk Mean Standar t, F
Deviasi

Ratio Ya Ya Ya Ya Mean Standar t, F

Deviasi
 UN-OBSERVED VARIABLE (LATENT VARIABLES)

Kepuasan
Kerja

Dimensi Job
Atasan Co
Worker

Indikator Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11

Q1 = setujukah bahwa pekerjaan saudara membosankan ? STS TS N S SS

Q4 = setujukah bahwa atasan saudara selalu menghargai pekerjaan saudara?
Q8 = setujukah bahwa teman sekerja sudara selalu mendorong saudara?
 Jenis variabel
• Variabel Independen (bebas)/exogen
• Variabel Dependen (terikat)/ endogen
• Variabel Kontrol (umumnya demografi)
• Variabel Moderating (memperkuat atau
memperlemah hubungan variabel
dependen dan independen)
• Variabel Intervening atau mediating
(variabel antara)
Reflective and Formative
Indicator

Stress Stress

Stomach Headache tension Divorce Retired Bankcrupt

ache

Reflective Indicator Formative Indicator

 Independent Dependent

CAR

ROA

NPL

NIM
Size Age

Control variables

ROA = b0 + b1CAR + b2NPL + b3NIM + e

ROA = b0 + b1CAR + b2NPL +b3NIM + b4Size + b5Age + e

We never hyphotese variable control

 Path Analysis

Exogenous variable ξ (Ksi)= is not posses an antecendent varaibles

Endogenous variable η (Eta)= is posses an antecedent variables
Intervening /Mediating variable = is posses an antecedent as well as
consequent variables (η1)
Coefficient parameter from Exogenous to endogenous γ (gamma)
Coefficient parameter from endogenous to endogeneus β (beta)

η1=γ1.1ξ1+γ1.2ξ2+ ζ1

η2= γ2.1ξ1+γ2.2ξ2+β2.1η1+ ζ2
 Structural Equation Modeling

Measurement Model: Structural Model:

ξ 1(Ksi)
X1=λ1ξ1+d1 η1=γ1.1ξ1+γ1.2ξ2+ ζ1
X2=λ2ξ1+d2
X3=λ3ξ1+d3 η2= γ2.1ξ1+γ2.2ξ2+β2.1η1+ ζ2
 Moderating Variables

X Y

Numbers of Capability of
Books read reading

Parent
Condition
illiterate
Z

Y = b0 +b1X + b2Z*X+ e

Y= b0 + b1X + b2Z +b3Z*X + e

 Universal vs Contingency Theory

A B

A and B as Shoes Company

If I have to setup MIS in company A, is it also applicable for company B
According Universal theory, yes it also can be applied as well at company B

But according to Contingency theory can not be applied to B since company B

Has its own situational or contextual factors that might be different from A,
The contingency variables are:
1. Environment uncertainty
2. Technology
3. Organizational structure
4. Strategy
5. Culture
Indepeden Dependen

Partisipasi Kinerja
Anggaran Manajerial

Dependen
Independen
Partispasi Kinerja
Anggaran Manajerial

Motivasi
Moderating Dependen
Independen
Partisipasi Kinerja
Anggaran Manajerial

Motivasi

Intervening
 Hypotheses
• A Special type of proposition that represents
conjecture about concepts stated in empirically
testable form.
• Is derived from a proposition by linking one or
more constitutively defined concepts with the
empirical indicators identified in the operational
definition.
• Proposisi: suatu pernytaan tentang konsep yang
dapat dinilai benar atau salah jika dikaitkan
dengan fenomena.
• Jika proposisi ini dirumuskan untuk pengujian
empiris, maka disebut hipotesis
 Hypotheses
• Hipotesis sering juga digambarkan sebagai suatu pernyataan dimana kita
mengkaitkan variabel kepada kasus. Kasus adalah entity atau sesuatu
yang sedang dibahas dalam hipotesis.
• Hipotesis dapat dibedakan menjadi
A. Hipotesis deskriptif (descriptive hypotheses): yaitu proposisi yang
menyatakan tentang keberadaan, ukuran, bentuk, atau distribusi suatu
variabel.
Kota-kota di Amerika mengalami kesulitan anggaran.
Kota-kota = kasus
Kesulitan anggaran= variabel

Delapan puluh persen pemegang saham PT XYZ menyukai kenaikan

deviden kas.
PT XYZ= kasus
Sikap pemegang saham yang menghendaki kenaikan deviden= variabel
 Hypotheses
• Beberapa peneliti lebih senang menggunakan pertanyaan penelitian
daripada hipotesis deskriptif.
Apakah kota-kota di Amerika mengalami kesulitan anggaran?
Apakah para pemegang saham PT XYZ menyukai kenaikan deviden
kas?
B. Hipotesis Relational : merupakan pernyataan yang menggambarkan
hubungan antara dua variabel dalam kaitannya dengan suatu kasus

Mobil impor dipandang memiliki kualitas yang lebih baik oleh

konsumen Indonesia dibandingkan dengan mobil buatan dalam
negeri.
Konsumen = kasus
Negara asal dan kualitas = variabel
 Hypotheses
• Hipotesis relational dapat berbentuk:
1. Hubungan korelasi: tdk ada hubungan sebab-akibat
Masinis muda (umur<35th) lebih rendah produktivitasnya
dibandingkan masinis yang berumur 35th atau lebih.
2. Hubungan sebab-akibat: dalam hal ini adanya
perubahan satu variabel akan mempengaruhi variabel
lainnya.
Adanya kenaikan pendapatan keluarga akan
meningkatkan prosentase pendapatan yang ditabung.
 Hypotheses
• Bentuk baku hipotesis
• Null Hypotheses (H0)
Tidak ada beda antara x dan y
Tidak ada asosiasi/korelasi antara x dan y
Tidak ada pengaruh x terhadap y
• Alternative Hypotheses (HA)
Ada beda antara x dan y
Ada asosiasi/korelasi antara x dan y
Ada pengaruh x terhadap y
• Hipotesis yang baik hrs memenuhi tiga hal:
a. Memenuhi tujuan penelitian
b. Dapat diuji
c. Lebih baik dari rivalnya.
The Impact of Structure, Environment, and Interdependence on The Perceived
Usefullness of management Accounting System
Robert H Chenhall and Degan Morris
The Accounting Review Vol LXI No 1 January 1986
External
Environment
Uncertainty
Moderating Perceived
Usefulness
Organizational Of MAS:
Structure: 1. Scope
Decentralization 2. Timeliness
3. Aggregation
4. Integration

Organizational
Independence
 Hipotesis Penelitian
• H1: There is a direct relationship between the
perceived usefulness of the MAS characteristic
of broad scope and the three contextual variable
of 1)decentralization, 2) perceived environment
uncertainty, 3) organizational interdependence.
• H2: There is direct relationship between the
perceived usefulness of the MAS characteristic
of timeliness and the contextual variable of PEU.
• H3: There is direct relationship between the
perceived usefulness of the MAS characteristic
of aggregation and the two contextual variables
of 1) decentralization, and 2) PEU
 Hipotesis Penelitian
• H4: There is direct relationship between the perceived
usefulness of the MAS characteristic of integration and
the two contextual variables of 1) decentralization, and
2) organizational interdependence.
• H5: There is an indirect relationship between the
perceived usefulness of the two MAS characteristic of 1)
broad scope and 2)aggregation, and PEU acting through
decentralization.
• H6: There is an indirect relationship between the
perceived usefulness of the two MAS characteristic of 1)
broad scope and 2) integration, and OI acting through
decentralization.
Influences on Departemental Budgeting: An Empirical Examination of
A Contingency Model

Kenneth A merchant
Accounting, Organizational and Society Vol 9 1984

Situational
Characteristics
-Production Budgeting
System Organizational
Technology Performance
-Market factors Characteristics
-Organizational
Characteristics
 Hipotesis Penelitian
• Production technolgy and budgeting
H1: Department with productive technology that are
relatively routine and repetitive tend to place greater
emphasis on formal budgeting.
• Market factors and budgeting
H2: Department with predominantly in the earlier stages of
product life cycle (emerging, growth) tend to place less
emphasis on formal budgeting than do departments with
products predominantly in the later stage (mature,
declining).
H3: Department with a relatively strong market position
tend to place greater emphasis on formal budgeting.
 Hipotesis Penelitian
• Organizational characteristics and budgeting.
H4: Larger, more diverse department tend to place
greater emphasis on formal budgeting.
• Effects on organizational performance
H5: Organizational performance tend to be higher
where there is a “fit” between the use of
budgeting and the situational factors as
described in Hypotheses 1 -4
The Effect of Technological Sophistication on Strategic Profiles,
Structure and Firm Performance

Jeffrey G Covin and E Prescot

Journal of Management Studies 27:5
September 1990

Industry Technological Sophistication

Macro Strategy Micro Strategy Organizatioanl Firm

Variable Variable structure Performance
 Hipotesis Penelitian
• H1: Small firms in high-tech industries have mission
strategies which are more build-oriented than are those
of small firms in low-tech industries.
• H2: Small firms in high-tech industries have strategic
postures which are more entrepreneurial than are those
of small firms in low-tech industries.
• H3: Small firms in high-tech industries have structures
which are more organic than are those small firms in
low-tech industries.
• H4: Small firms in high-tech industries perform better
than do small firms in low-tech industries.
 Diversification and Performance: Critical Review and Future Directions
Deepak K Datta and Nandini Rajagopalan
Journal of Management Studies 28:5 September 1991
Degree of Diversification Mode of
diversification strategy diversification

Industry structure
-Concentration
-Barrier to entry
-Capital intensity
Moderating
-Growth

Organizational
Factor:
Structure,culture
system
Performance:
-return
-risk
- other
 The Effect of Effort on Sales Performance and Job Satisfaction
By Steven P. Brown and Robert A Peterson
Journal of Marketing Vol 58, April 1994

Competitiveness Instrumentality
Role ambiguity Role Conflict

Effort

Performance Satisfaction
The Design and Effects of Contriol System : Test of Indirect Models
By: Michael D Shields; F Johny Deng and Yutaka kato
Accounting, Organization and Society 25 2000

Participative
Std setting

Std Std Based

Tightness Incentive

Job Related
Stress

Job
Performance
 Setting Penelitian

Laboratory Setting Field Setting

Laboratory Research Field Research

Laboratory Experiment Field Field

Experiment Study
 Common Type of Dependent Measure of Variables in
Accounting Research

Behavioral Measures Self Report Measure

Frequency Latency Number of Rating Scale

Of respond errors

Research involves Likert Scale

Measurement Attitude measurement

Established measures Develop own measure

Through literture review

Validity and Reliability Need to test for validity

Known and reliability
 Desain Penelitian
• Non-Experimental • Experimental Design
Quasi-Experimental Based on experimental
Design research strategy
Based on correlational - Simple experiment
research strategy - Multiple group
- Observational research experiment
- Case study research - Field experiment
- Content analysis
- Survey research
- Quasi-experimental
• Quasi experimental design if the data come from an environment which existed or
Or events that occurred without his direct intervention.
• True experiment design if he is (1) able to manipulate explicitly one or more inde
pendent variables, (2) to assign subject to control or experimental groups
(basically to achieve randomization)
 Sampling Method
• Unit analisis:
1. individuals groups artifacts (books,
photos, newspapers)
2. geographical units (town, census tract,
state) .
3. social interactions (dyadic
relations, divorces, arrests)
 Probability Sampling

Probabilitas

Populasi
Persh Go Public Sample
Di BEJ 60 persh
350 persh

Mean

Mean Std. Dev

Variance
Std Dev Variance
 Probability Sampling
• Simple random sampling:
1. Lotere
2. Tabel Random
• Complex Random Sampling:
1. Systematic random sampling
2. Stratified random sampling
3. Cluster random sampling
 Non-Probability Sampling
• Convenience sampling
• Purposive sampling
• Judgemnetal sampling
Kesalahan Sampling:
Type I error (alpha): salah padahal benar
Type II error (beta): benar padahal salah
 Menentukan Besarnya Sample
• Ada 3 faktor yang mempengaruhi sample size:
a. Variance atau heteroginitas dari populasi (S)
b. Besarnya acceptable error (E)
c. Confidence level (Zc.i)
• Menentukan besarnya sample:
a. Estimate std dev populasi (kasar 1/6 dari
range)
b. Lakjukan judgment tdp kesalahan yang
diinginkan.
c. Tentukan confidence intervalnya
 Perhitungan besarnya sample
• Rumus : n= (ZS/E)2
Dimana:
Z = nilai cofidence level
S = std dev. Sample atau populasi
E = nilai accepted error, plus atau minus
error factor (one half total confidence
interval)
 Perhitungan besarnya sample
• Seorang peneiliti ingin meneliti
pengeluaran untuk lipstick dan
menginginkan besarnya confidence level
95% (Z) dengan range kesalahan (E)
kurang dari $2 dengan standar deviasi
$29
n = (ZS/E)2 = [(1.96)(29)/$2)]2= 808, jika
error range menjadi $4 ,maka sample
n = [(1.96)(29)/$4)]2 = 202
 Uji Kualitas Data
• Uji Reliabilitas atau konsistensi : cronbach
alpha
• Uji validitas :
a. Bivariate korelasi
b. Analisis faktor
• Uji Non response bias
 Metode Analisis Data
• Metode Dependen: menguji ada tidaknya
hubungan dua set variabel. Satu variabel adalah
variabel bebas (independen) dan satu variabel
adalah variabel terikat (dependen). Tujuan
metode dependen menentukan apakah variabel
bebas mempengaruhi variabel terikat.
• Metode Interdependen:jika set data yang ada
tidak mungkin dikelompokkan kedalam variabel
bebas dan terikat, maka yang dapat dilakukan
adalah mengidentifikasi bagaimana dan
mengapa variabel tersebut saling berkaitan satu
sama lain.
 Teknik Analisis Data
Variabel terikat Variabel Bebas Teknik Analisis

1. Satu metrik Satu non-metrik Uji Beda t-test

Dua kategori

2. Satu metrik Satu non metrik Analysis of variance

>dua kategori (ANOVA) –one way
Anova
3. Satu metrik Dua non metrik Two way Anova
>Dua kategori

4. Dua atau lebih Satu atau lebih non Multivariate analysis

metrik metrik of variance (Manova)
 Teknik Analisis Data
Variabel terikat Variabel bebas Teknik Analisis

5. Satu metrik Dua atau lebih Multiple Regression

metrik dan non
metrik
6. Satu non metrik Analisis Diskriminan
Dua atau lebih Logistic regression
metrik dan non
metrik
7. Dua atau lebih Dua atau lebih Structural equation
metrik metrik Modelling (SEM)