Methodology Writing

Methods
• Research design
• Setting
• Developing intervention protocols
• Sample and Sampling
• Data collection methods
• Ethical consideration
• Data analysis plan
Design

•The overall plan for addressing a research

problem
• Broad design options and intervention :
• Experimental
• Quasi-experimental
• Nonexperimental (observational study):
Experimental Designs
•Posttest-only (or after-only) design
•Pretest–posttest (before–after) design
•Crossover design
•Factorial design
 Quasi-Experiments
•Nonequivalent control group designs
•1. Nonequivalent control group before–after
design
•2. Nonequivalent control group after-only design
•Within-subjects designs
•One-group before–after designs
•Time series designs
Nonexperimental Research
•1. Correlational Designs
•a. Retrospective Designs
•case–control design
•b. Prospective Designs (cohort study )
•2. Descriptive Research
•descriptive correlational
Research Design and Time
•Cross-sectional design
•Longitudinal design
•Trend studies
•Panel studies
•Follow-up studies
Continuum of designs for inferring causality

Strongest Weakest
Design Non Experimental Designs Design

______________________________________________________________________________
_
Other Research Designs:
1. Survey
2. Methodological research
3. Case study
4. Secondary analysis
5. Meta-analysis and meta-synthesis
6. Evaluation research
Example: Design
•A randomized controlled trial, with
pre-, post-test design with two
groups (intervention and control
group) will be used.
Setting:
•The specific location where
information is gathered.
Setting: example
• The study sample was derived from an antenatal
clinic of a major, large, teaching, maternity, referral
hospital in the North of Jordan. This hospital is
controlled by the government and directed by the
Ministry of Health, and it provided health care
services free of charge for all pregnant women. Nearly
2,500 women visited its antenatal clinics each month,
and around 30,000 women through all of 2015.
Sampling Plans
• Target population
• Accessible population
• Sampling
• Eligibility criteria
• Inclusion criteria:
• Exclusion criteria:

• Sampling Problems in Quantitative Research

• Sampling bias
• Sampling error
Populations

Elsevier items and derived items © 2009, 2005, 2001, 1997 by Saunders, an imprint of Elsevier Inc.
14
Sampling Goal in Quantitative Research
•Representative sample

•More easily achieved with:

•Probability sampling
•Homogeneous populations
•Larger samples
Types of Sampling Designs

•Probability sampling

•Nonprobability sampling
Types of Nonprobability Sampling
1. Convenience sampling
2. Snowball (network) sampling
3. Quota sampling
4. Purposive sampling
5. Consecutive sampling
Types of Probability Sampling

•Simple random sampling

•Stratified random sampling
•Cluster (multistage) sampling
•Systematic sampling
 Sample Size calculations
•The number of study participants in the final
sample
•Sample size adequacy is a key determinant of
sample quality
•Sample size should be estimated through
power analysis.
• The larger the group difference is, the smaller the
sample will be
 Steps in Sampling
1. Identify the population. Researchers often begin by identifying
an accessible population, and then decide how best to define the
target population.
2. Specify the eligibility criteria.
3. Specify the sampling plan.
(a) the method of drawing the sample and
(b) how large it will be.
4. Sample Recruitment: (Face-to-face , online,..)
5. Refusal and response rate
Sample and Sampling:
Example
Target and accessible: example
•The target population for this study was
anemic pregnant women
•the accessible population was the
anemic pregnant women who were in
the antenatal clinic in Northern Jordan
Inclusion and exclusion: example
• The inclusion criteria were: pregnant women
diagnosed with anemia, aged 18-45 years, in their
second trimester of pregnancy from 13 to 28
gestational weeks, with hemoglobin levels below
10.5 mg/dl, and who had smart phones with access
to social media (WhatsApp application)
• Pregnant women who were illiterate or had a history
of obstetric, medical or psychological problems were
excluded.
CONSORT: example
• The total number of women approached for screening
for eligibility was 4969 women;
• the proportion of those who were screened for
eligibility was 99% (n=4919);
• the proportion of screened women who were found
to be eligible was 4% (n=216),
• and the participation rate among those found to be
eligible was 96% (n=208) (Figure 1).
sample size calculation: example
• Based on a power analysis,
• with a two-sided α of 0.05,
• a beta of 0.20
• and a medium effect size of 0. 5 (Cohen, 1988),
• the total sample size was calculated to be 128 anemic pregnant
women.
• We increased the sample size up to 200 participants to allow for
dropouts and potentially to enhance the statistical power of the
study.
Sampling techniques: example
• Random assignment was used to divide the women into one of two
groups.
• Each participant was given a number from 1 to 200 and randomly
assigned into one of the two groups with the aid of a computer-
generated table of random numbers.
Attrition: example
• Eight participants withdrew from the study for different reasons:
• two of them refused to complete the post-test questionnaire;
• two participants were delivered preterm,
• and we lost the communication with four participants.
Intervention protocol: example
• Health Information Package Program (HIPP) was used to educate
anemic women.
• The adopted draft was translated from …
• video of 15 minutes duration
• PowerPoint slides
• The areas covered in the video records were…..
• The video and its content were validated by specialists
• The participants in the control group received standard care in
antenatal clinics,
Major Types of Data Collection
Methods

•Self-reports: by questioning people

•Observation: observation of people’s
behavior
•Biophysiologic measures: to assess
important clinical variables
 Structured Self-Reports
•formal instrument.
•Interview schedule
•asked orally.
•Questionnaire
•self-administered by respondents
Types of Questions in a Structured
Instrument
•Closed-ended
•Open-ended questions
Some Types of Close-ended Questions

1.Dichotomous items: yes/no

2.Multiple- choice questions
4.Rank order questions:
5.Visual analogue scales
6.Composite Psychosocial Scales:
Likert scales
32
Response
• Examples:
Set Biases
• Social desirability response set bias:
• Extreme response set:
• Acquiescence response set (yea- sayers)
• Nay-sayers response set
• To decrease such biases, researcher should
counterbalance the statements (positive
and negative statements)
Observations
•1. Category systems  Checklists
•2. Observational Rating Scales
Instruments: example
• A structured interview questionnaire was used for
collecting data.
• This instrument was developed by….
• Its Cronbach alpha was 0.938
• The questionnaire consisted of four parts.
•The first part of the questionnaire
consisted of two sections:
• section one asked about socio-demographic data,
• and the second section asked about health-related
information.
•
Instruments: example
• The second part of the questionnaire was about
• The third part of the questionnaire was about food selection ability,
which consisted of three checklists asking participants to identify food
items which were rich and poor sources of iron (17 items), protein (21
items) and vitamin C (15 items). Out of these 53 items only 30 were
correct.
• The total score of food selection ability was 30.
• The fourth part of the questionnaire was the Structured Knowledge
Interview Schedule (SKIS). Each correct answer was assigned one
point.
Validation of Instruments: example
• The instrument was translated from English to Arabic and
translated back from Arabic to English by experts.
• Content validity index CVI was 0.91, which was considered
highly valid.
• A Cronbach’s alpha for the Arabic version of the adopted
instrument was 0.938.
• A Pilot study was conducted on 20 anemic pregnant women
(10%) to test the tool content, applicability, clarity and time
needed to complete the interviewing questionnaire.
Data Collection Procedure
• researcher scheduled a meeting with the director of the hospital
• use a private room in the antenatal clinics
• A team of three trained
• were responsible for identifying and inviting mothers who met the
eligible criteria and were willing to participate in the study.
• took approximately 10-15 minutes to be completed
• The post-test four weeks after recruitment
• data collection day for three months from the 1st of April 2016 to 4th
July 2016.
Ethical Principles: The Belmont Report

• Beneficence
• Respect for Human Dignity
• Justice
 Procedures for Protecting Study Participants
1. Risk-benefit assessments
2. Informed consent
3. Confidentiality procedures
4. Debriefings (ask questions )and referrals
5. Special treatment of vulnerable groups
6. Institutional Review Boards
Ethical Considerations: example
•Ethical approvals from the Institutional Review
Board at Jordan University of Science and
Technology and the Ministry of Health were
obtained
•Signed informed consent was obtained from all
participants
•The researcher assured the participants that no
harmful effects
Ethical Considerations: example
•The researcher explained to the participants that
their participation would assist with the
development of a new strategy
•assured that their privacy as well as the
confidentiality of their response was protected.
•their participation in this study was voluntary
and that they could withdraw from the study at
any time, without bearing any consequences.
Statistical Analysis
1. Descriptive statistics
2. Inferential statistics
Statistics: Descriptive
1. Univaraite
1. “frequency distribution”.
2. Measures of central tendency (mode,
median, mean), and
3. measures of variability (standard deviation
, variance, range).
2. Bivaraite analysis:
1. contingency table,
2. correlation- such as:
1. Pearson or Product Moment Correlation
Coefficient, and
2. Spearman’s correlation coefficient).
 Inferential Statistics
The most common used inferential statistics are:
1. t-test,
2. ANOVA,
3. Chi-square,
4. Mann-Whitney U,
5. correlation coefficient,
6. simple linear regression,
7. multiple regression,
8. ANCOVA,
9. MANOVA, and
10. path analysis.
Hypothesis Testing
• Based on rules of negative inference:
• research hypotheses are supported if null
hypotheses can be rejected.
• Involves statistical decision-making to either:
1. accept the null hypothesis (no relation ship)
or
2. reject the null hypothesis (there is a
relationship)
 Hypothesis Testing
• 1. Researchers compute a test statistic with their data
(using appropriate formulas ) and
• 2. then determine whether the statistic falls beyond
the critical region in the relevant theoretical
distribution.
• Values beyond the critical region indicate that the
null hypothesis is improbable (impossible), at a
specified probability level.
critical region
Data Analysis: example
• The data were analyzed using the Statistical Package For Social
Sciences, SPSS, for Windows, version 20.
• Descriptive statistics were used to demonstrate demographic
characteristics
• A Chi-square analysis was used to compare the study groups for
categorical demographic variables.
• Analysis of covariance (ANCOVA) was performed to identify potential
effects of the health information package program
• A p value of 0.05 or less was considered as statistically significant