Dr James Betts FACSM

J.Betts@bath.ac.uk
@DrBSteamjets
Lecture Outline:
•The Research Process
•The Research Design Continuum
•Experimental Designs
•Sampling Methods
•Scientific Reasoning
•Quantitative & Qualitative Research Strategies.
 What is Research?

1. Systematic – research process

2. Logical – induction/deduction
3. Empirical – evidence based
4. Reductive – generalisation
5. Replicable – methodology.
 Review the Available
Literature

Publish Formulate a
Findings Question
Research Process
Interpret Select an Appropriate
Findings Research Design

Collect Relevant
Data
Research Continuum

Reductionism
 Research Continuum
Basic Applied
ØTheoretical? ØQuick Answers?
ØMore Invasive? ØLess Invasive?
ØLaboratory Based? ØField Based?
ØTightly Controlled? ØLoosely Controlled?
Internal Validity?
ØLacks External Validity? > ØExternally Valid?
ØFocus on Mechanism ØFocus on Effect
ØMore Reductionist ØLess Reductionist.
 Research Continuum
Basic Applied
e.g.
ØDoes Caffeine Ingestion
Improve Athletic
Performance?
 Research Continuum
Basic Applied
e.g. e.g.
ØDoes Caffeine Ingestion ØDoes Caffeine Ingestion
Improve Ca2+ binding Improve Athletic
with troponin? Performance?
-Would this Facilitate
Acto-Myosin Coupling?
-Would this aid
contraction?
 Research Continuum
Basic Applied
e.g. e.g.
ØDoes Caffeine Ingestion ØDoes Caffeine Ingestion
Inhibit Glycogen Improve Athletic
Phosphorylase? Performance?
ØDoes Caffeine Ingestion
Increase Lipid
Metabolism?
-Would this Spare
Endogenous Glycogen?
 Research Continuum
Basic Applied
e.g. e.g.

ØDoes Caffeine ØDoes Caffeine Ingestion

Ingestion Stimulate the Improve Athletic
CNS? Performance?

-Would this Increase

Motor Unit Recruitment
-Would this Reduce
Perceived Effort?
 Research Design Continuum
Research Design

Analytical Research Experimental Research

Descriptive Research Pre-designs

Reviews
Quasi-designs
Philosophical Case Study Survey

Historical True-designs
Cross-Sectional
Meta-Analyses
Longitudinal
Statistical-
Correlational designs
 Analytical Research
• Reviews
– A critical account of present understanding
– A meta-analysis is a quantitative method of review
• Historical Research
– Accessing both primary (e.g. witnesses) or secondary
(e.g. literature) sources to document past events
• Philosophical Research
– Organising existing evidence into a comprehensive
theoretical model
 Descriptive Research
• Case Study
– Accrual of detailed information from an individual

• Survey Refutable?
– Cross-sectional: Status of a various groups at a given
point in time
– Longitudinal: Status of a given group at various points
in time
– Correlational: Relationships between variables
 Correlational Evidence
• When variable X increases, variable Y also
increases
• So, does X increase Y?
– or does Y increase X?
• Alternatively, does Z increase both X and Y?

Correlations do not infer Causality

(and vice versa?)
See inapt use of language: Brown et al (2013)
i.e. always Read Primary Lit (inc. actual data)!
 Correlation r=0.87

Correlation r=0.81

http://t.co/vWOyN0N1IB
 Correlation r=0.-83

Correlation r=-0.98

http://t.co/vWOyN0N1IB
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Independent Variable = this variable is the ‘cause’
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Independent Variable = can be manipulated or allowed to vary
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Independent Variable = also known as the predictor variable
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Dependent Variable = this variable is the ‘effect’
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Dependent Variable = should only vary in response to the IV
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Dependent Variable = also known as the criterion variable
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Law of the single variable:
there will always be uncontrollable influences
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Extraneous Variables = must be controlled to isolate the
effect of the IV on the DV
 Experimental Research
• Experimental research involves a direct assessment of
how one variable influences another
• This allows the establishment of causality
• All extraneous variables must be held constant while a
single variable is manipulated and the effect measured

• Definition of variables:
Confounding Variables = extraneous variables which have
co-varied with the IV
 Experimental Designs
• Pre-Experimental
• Quasi-Experimental
• True-Experimental

Key:
– R = random assignment for equivalent groups
 Random Group Assignment
List 20 individuals
All to be assigned to treatment (T) or placebo (P)

Group 1: toss a coin for each individual

Group 2: ‘think-up’ a list that seems random.

 Experimental Designs
• Pre-Experimental
• Quasi-Experimental
• True-Experimental …or via repeated measures design,
matched pairs design or
matched groups design
Key:
– R = random assignment for equivalent groups
– O1,2… = observation of group x (recording of DV)
– Oa,b… = observation of group y (recording of DV)
– T = treatment (IV)
– P = placebo (IV).
 Experimental Designs
• Pre-Experimental
• Quasi-Experimental
• True-Experimental

Question:

“Does protein supplementation

increase muscle hypertrophy?”
Pre-Experimental Designs
One Shot Study

T O 1
 Pre-Experimental Designs
One Group Pre-test Post-test

O
1 T O
2
Pre-Experimental Designs
Static Group Comparison
T O 1

P
Oa
 Pre-Experimental Designs
Static Group Comparison
O 1

Oa

Daniel 1:8
 Quasi-Experimental Designs
Time series

O
1 O2 O 3 T O
4 O
5 O
6
 True-Experimental Designs
Randomised Group Comparison
T O 1

R
P
O 2

Earliest recorded example of random group

allocation as recent as 1928 (Forsetlund et al. 2007)
 True-Experimental Designs
Pre-test Post-test Randomised Group Comparison
O 1 T O 2

R
O 3
P O 4
 True-Experimental Designs
Solomon Four-Group Design
T
O
1 O
2

P O
O 4

R
3

T
O
5

P O
6
 Sampling
-Split into research teams
-Each person take a ‘sample’ of Smarties

-Each group record the total number of

Smarties and the number of red Smarties
 Sampling
Target
Pop.
(N)

Sample (n)
• Effective Sampling produces a n which is representative of N
• Note: n is only ever representative of the N it was drawn from,
i.e. not necessarily the general population.
 Sampling

Statistics

The dependent variable can be

generalised from n to N
 Sampling Methods
• Random- All members of N have an equal
chance of selection
e.g.

• Stage- Randomly select a group, then take

sample School
e.g. Class

• Cluster- Select a natural group to sample from

e.g. local community
 Sampling Methods
• Stratified- identify strata and sample accordingly
i.e. Global Pop. Sample (n=100)
= 51% = 51
= 49% = 49
• Systematic- e.g. every fourth person but starting
at a random point

• Opportunity- sample a convenient group

Avoid Researchers!
 Scientific Reasoning (Logic)
Quantitative?
Confirmation of a theory from
your own observations

General Deductive Reasoning Specific

Theory Inductive Reasoning Observation
Formation of a theory grounded
in your own observations
Qualitative?
 Quantitative versus Qualitative
Quantitative Research Strategy Qualitative Research Strategy
•Investigation aims to assess a pre- •Investigation aims to create a novel
stated theory (Deductive Reasoning) theory (Inductive Reasoning)
•Often involves hypothesis testing •Researcher becomes an inherent part
of the study - ethnography
•Attempts to minimise the influence
of the researcher on the outcome •Qualitative data infers complex
statements or opinions
•Quantitative data infers statistics
•Data collection therefore permits
•Data collection therefore requires
‘open’ responses
‘closed’ responses
 Choice of Research Strategy…
• Based on:
– Epistemology (How should we be attempting to assess
knowledge?)
• Positivism = explain a phenomena
• Interpretivism = understand a phenomena
– Ontology (Does the data exist in a tangible or an intangible
form?)
• Objectivism = explain independent external outcomes
• Constructionism = understand how social factors interact
 Choice of Research Strategy…

• Study in the natural sciences often requires a

positivistic epistemology and an objectivistic
ontology
• Study in the social sciences often requires an
interpretive epistemology and a constructionist
ontology
• However, it is occasionally possible to combine
these strategies by coding qualitative data
quantitatively (i.e. Athlete = 1 ; Non-Athlete = 2)
 Selected Reading

• Thomas J. R. & Nelson J. K. (2005) Research

Methods in Physical Activity, 5th edition.
Champaign, Illinois: Human Kinetics

• Berg K. E. & Latin R. W. (2008) Essentials of

Research Methods in Health, Physical Eduction,
Exercise Science, and Recreation, 3rd edition.
Maryland: Lippincott Williams &Wilkins