FNR201 - Lecture cards

What is research

• systematic process of collecting and analyzing informaiton

• process of discovering or uncovering new knowledge

  • fundamental basic research principles

• encomapsses facts or understanding that are not ends in themselves but are compontnets in total process

• based on assumptions

Good research is

• based on the work of others

  • identify current state of knowledge

  • identify knowledge gaps

  • research agenda

• activity that can be replicated

  • this is a mark of credible research

  • can help to form a basis for further work in the same area

• generalizable to other settings

  • stand up in different but related settings

  • some may be difficult to generalize

• based in logical rationale and tied to theory

  • fit into bigger pitcure

  • places findings into current state of knowledge

• Doable

  • feasible (not to complicated or expensive to actually run

  • clear questions with a clear timeframe

• able to generate new questions

  • give oppurtunity for further studies

• Incremental

  • makes small contribution to larger body of knowledge

why do research?

• improves understanding

• allows for interpretation

• determines the strength of relationships between two variables

• allows for in-depth analysis (qualitative)

4 goals of research

• description

• explanation

• prediction

• control

Methodology

• set of rules and procedures for how research is to be conducted

• this is the experimental design of a study

methods

• specific data collection and analysis techniques

  • analysis procedures and techniques for gathering the evidence to be used

• this is more specific about what is used and what is being measured

Positivist approach to research

• emphasis on control of physical and social environments

• want results to operate like the natural world

• used in physical sciences

• quantitative and experimental knowledge

• uses deduction

  • taking questions and finding answers

• Instrumental approach

Positive approach assumptions

• all behaviour is naturally determines

• humans are part of the natural world

• nature is orderly and regular

• all objective phenomena are eventually knowable

• nothing is self - evident

• truth is relative

• knowledge comes from experience

Interpretive approach

• aims to examine how people make sense of their lives and how the social world operates

• naturalistic qualitative knowledge

• emphasized that humans are always in the process of becoming

• values are relative and thus change over time

• concerned with issues of dependability and transferability

critical approach

• examines societal structures and power relations present in our world

• based on reflective knowledge

• values are absolute

  • certain values are correct while others are not

• assumes all behaviour is influenced by power issues

• goal is to achieve or promote social justice

Descriptive research

• what is happening

  • census

  • public opinion polls

  • market research

Explanatory research

• why does x change y

  • why does soluble fibre intake (x) lower cholesterol levels (y)?

Pure/basic research

• goal to expand knowledge and test theories

• values knowledge for its own sake

• not trying to promote particular theories

• value free research

Applied research

• goal of the research is to solve a problem

• changing health care practices

• focuses on variables that can be changed by intervention to achieve desired goals

Quantitative research

• quantify observations about human nature

• emphasis on precice measurements and the testing of hypothesis and data analysis

• single reality

  • large samples

  • deductive (hypothesis testing)

  • objective reality

  • tests theory

  • outcome oriented

  • positivist

  • statistical analysis

Quanlitative research

• understanding how participants experience and explain their own world

• ephasiz on verbal descriptions and explinations of human behaviour

  • multiple realities

  • small samples

  • inductive (observations → hypothesis)

  • reality is socially constructed

  • develops theory

  • process oriented

  • rich descriptions

  • control is not important

  • interpretive results

  • critical analysis

questions asked in quantitative research

• what is the relationship between obesity and tv viewing

• what is the relationship between breakfast programs and cognitive funciton

• what is the relationship between exercise and fat mass

qualitative questions asked

• what is is like to be overweight and obsese

• how do teachers deal with students who do not follow instructions

• what is it like for international students to study abroad

Deduction

• top down approach

• concerned with testing hypothesis

• quantitative

• theoretical → conceptual → operational

Induction

• bottom up approach

• exploratory

• qualitative

• operational → conceptual → theoretical

Units of analysis

• individuals

  • focus of surveys and experiemnts

  • individuallevel variables

• Aggregates

  • properties of groups (a population)

  • differences in health in different ethnic/national groups (epidemiology)

  • may not be able to apply results to each individual but this is the distribution of scores in general

What data is at the very bottom of the pyramid

• expert opinions

• no one cares that you are en expert

Dependent variable

• effect of cause and effect relationship

• influenced by other variabeles

Independent variable

• influences other variables

• cause in cause and effect

Control variable

• variable that could influence the releationship between independent and dependent

• intervening, conditional and confounding

• grouped seperately cause it might have alternate effect

Intervening variable

• variable that links an independent and dependent variable

• represents an explination of how the independent variable influences the dependent variable

  • on the causal pathway

conditional variable

• variable that accounts fo rthe change in the relationship between independent and dependent when the general conditions change

Confounding variable

• has a possible influence on both the independent x and dependent variable y

  • accounts for the relationship between x and y

• affects the two variables independently

  • can make it seem like there is a relationship where there is none in reality

defining dependent and independent variables

• dependent

  • left alone

  • measured

  • varied at unknowl rate

  • effect

• Independent

  • intentionally manipulated

  • controlled varied at known rate

  • cause

Operationalization

• key is that variables must be measurable

  • reliable

    • repeatable

  • valid

    • are you measuring what you think you are measuring

Example for operalization ***

• seniors appetite

• should not measure meals consumed per day

  • cannot get valid measures because there is no setpoint for measuring meals vs appetite

Research process

• identify a topic of interest

• gather background information

• set research questions

• choose a design and methodology

• develop necessary tools and methods

• access sample to collect data

• organize and analyze data

• interpret results and define conclusions

• transfer knowledge

planning research questions qual vs quant

• quantitative

  • hypothesis and objectives

• qualitative

  • aims and objectives

parts of planning research questions

• aim: broad statement of desired outcome

• hypothesis: propositon about the relationship between the tested variables

• objective: steps necessary to answer the research question

What is epidemiology

• research discipline involved with distribution and determinants of health related states or events in populations

  • distribution of health related determinants of health and disease

  • how often related events occur in diff groups of people

  • why variations in patterns of health and disease exist in different populations

  • application of the analysis to control health issues in populations

Epidemiology word

• study upon populations

  • epi (among)

  • Demos (people)

  • Ology (study)

epidemiology importance

• understanding the cause of disease in populations

• prevention and control of disease in pops

• guide health care policy and planning (policy makers)

• guide management of health and disease in individuals

uses of epidemiology

• monitor health care use

• assess pop requirements for health care

• evaluating organization of health are services

• improving patientcare

• generation hypothesis

• monitoring regionla variations in provision of services

• investigating the impact of inequality

population

• large groups of ppl in defined setting

• relatively unselected group of ppl

  • general population, hospitalzed poplation, uni pop

• subpopulation - shares specific characteristics (age, race, presence of disease etc)

sample

• research (epidimiologic research, clinical research) carried out in samples

• selected subset of the population

• representative samples

Epidemiology

• concerned about populations

• intersted in characteristics of a defined population

• for practical reasons it is done in samples

two questions of epidemiological research

• are the conclusions of the research corrects for people in the sample?

• if conclusions are true, does the sample fairly represent the population of interest?

  • generalizability: can these findings be applied to the broader population of interest?

3 explanations of research findings

• bias (systematic error)

• chance (random error)

• truth (observation is correct)

Bias (systematic error)

• process at any stage of the research tending to produce results that depart systematically from true values

• any stage of the research process can be susceptible to this error

• many types of biases (selection bias, measurement bias, confounding bias, outcome assessor bias)

  • qual will have more bias typically

  • less generalizable

Examples of Bias

• participants (mothers postpartum)

• intervention: higher intensity exercise (a) vs lower intensity exercise (b)

• outcome: weight loss at 6 months

• finding: women A more successful than those in B for weight los

• If A was healthier participants and b was less healthy

• if there was different conditions in the gyms

Descriptive epidemiology

• examines patterns and trends

• studying populations wihtout trying to change them

• reliance on existing data

• uses surveys of large groups of people to collect information

  • CCHS

  • NPHS

  • NHANES - us

Descriptive studies

• population characteristics

  • age, gender, race, SES etc

• Population trends

  • epidemic of childhood obesity

    • trend increasing

• Geographic factors

  • international comparisons

  • rural-urban comparisons

  • local comparisons

• Temporal trends

  • longterm trends (increasing childhood obesity)

  • seasonal variations (distribution of suicide with rates peakingin february

  • cyclic variation - time series analysis of seasonal variations

  • point epidemics

• Frequency of events

  • incidence and prevalence

point epidemic

• an epidemic in which several cases of a disease occur within a few days or hours due to exposure to a common source of infection

  • legionnaires disease

    • mystery illnness killing seniors in nursing home (env contamination) which got into the nursing homes water system

Incidence

• rate at which new events occur in a population during a specified period

  • number of new events in a defined population over a specified time / number of disease free people in a population at the strt of the time period

  • New cases/ population of area

Prevalence

• number of instances of a given disease or other condition in a given population at a designated time

  • number of existing cases in a defined population at a given time / number of people in the defined population at the same point in time

point prevalence

• measured at a single point in time for each person

  • eg. rate of obesity in one school at one measuring time

period prevalence

• measured during a specified period of time

Correlational studies (descriptive)

• compares disease frequency between different groups

  • mortality/morbidity rates in different countries

  • does not explain relationships or determine associations

cross-sectional studies

• describes the frequency of particular attribute in a defines population or sample at a given point in time

• provides a snapshot of the health experience for that population

• measurement of prevalence or distribution of an attribute

• contrasting prevalence rates can identify which populations have higher risk of disease

• helpful in measuring secular trends

• used as a basis for longterm followups

advantages and limits of cross sectional

• advantage

  • quick and easy

  • can measure multiple exposures and outcomes at one time

  • useful for measuring the burden of disease

• Limitations

  • problems with the direction of causality

  • recall bias

  • not efficient for rare diseases

  • not suitable for diseases with short durations

  • low response rates could bias prevalence estimates

Case series

• describe groups with similar conditions

• could describe postpartum obsessive compulsive disorder

• usually suggests future study hypotheses

Analytical epidemiology

• causation or etiology

• to test hypothesized cause and effect relationship between suspected risk factor and a disease or condition

analytical research designs

• observational

  • case control

  • cohort

• experimental

  • clinical trials

case-control studies

• group of patients who have particular outcome and a control group who do not have the outcome are compared

• measure the frequency of an exposure in both groups to see if it is more or less common in one group than the other

steps for case control studies

• define objective criteria for case diagnosis → selecting cases and controls → estimating exposure status

  • cases: clinical trials, hospital records

  • controls: population controls, hospital controls, relatives/friends

Advantages and disadvantages of case control

• advantages

  • relatively quick

  • more efficient for rare diseases than cross sectional studies

  • can test multiple exposure hypothesis on single disease

• Disadvantages

  • temporal sequence between exposure and outcome may be difficult to establish

    • which came first?

  • may be hard to obtain information about past exposures

  • selection of appropriate group may be difficult

Cohort studies

• follow a designated group of individuals for a given period of time

• study group is measured at baseline for characteristics then followed over tiem to assess the dvelopment of new events

• study health related events incidence in populations

  • explore and generation of hypothesses

  • examination of natural prognosis of disease

Cohort studies advantages and disadvantages

• advantages

  • direct measure of incidence of disease in exposed and unexposed groups

  • time relationship between exposure and disease can be studied

  • investigation of multiple outcomes can occur

  • natural history of disease can be investigated

• Disadvantages

  • time consuming

  • expensive

  • outcome assessment can be influenced by knowledge exposure

    • information bias

  • too much dependence on quality of records for historical cohorts

Clinical trials

• begin with 2 or more experimental conditions

• follow up of groups to specified endpoints or time points

• common measure: HR (hazards ratio)

Common measures in analytical epidemiological studies

• RR (cohort studies)

• OR (case control studies)

RR - Math

• incidence of cases in the exposed /

  incidence of cases in the unexposed

• A/(A+B) /

  C/ (C+D)

• Number of new cases in those expoed vs those not exposed

RR description

• measures the odds of an outcome occuring in one group compared to another

• ratio of probabilities

• used in cohort studeis to measure incidence

• How much more (or less) likely is the oucome in group A compared to group B

Odds ratio definition

• odds of an outcome occuring in one group compared to another

• a ratio of odds

• used in case control studies when the outcome is rare

• what are the odds of the outcome in group A relative to group B

Interpretation of odds ratio

• way to compare whether the probability of a certian event is the same for two groups

• an odds ratio of 1 implies thta the event is equally likely in both groups

• if it is above one it means that the event is more likely in the first group

• <1 means that the event is less likely in the first group compared to the second group

OR math

• odds of exposure for cases /

  odds of exposure for controls

• A/C /

  B/D

• AD/

  BC

Critical appraisal

• discipline for increasing the effectiveness of your reading

• useful tools for

  • systematically evaluating scientific literature

  • sifting the good from the bad when results are conflicting

  • breaking down barriers between research and practice

  • filtering out original research

critical apprasal of quantitative research

• ask yourself about

  • message

    • what were findings

    • what is the bottom line of this research

  • Validity

    • did trial address clearly focussed issue or question

    • how was sample selected? power analysis?

    • were groups similar at baseline?

  • Generalizability

    • can these be applied to a laeger populaitons?

    • were all possible outcomes considered?

    • do benefits outweigh th ecosts/harms?

• also condiser the declaration of outside interests, ethics and use of literature review

Appraisal of reviews

• message

  • does review set out to answer a specific question

• validity

  • have authors sought out the correct studies

  • was there explicit inclusion and exclusion criteria

  • how were results presented

• generalizability

  • if results are consistent across studies (even with different demographics) then results are generalizable

Qualitative study appraisal

• authors must demonstrate

  • clear study aims

  • choice of appropriate methodology

  • justification for who was included and who was not

• what is relationship between researcher and researched

Non-reactive studies

• participant is unaware that they are beign studies

• has no opportunity to react to researcher

  • use of information from stats canada to emasure food insecurity

Non-reactive study research designs

• comparative studies

• secondary data analysis

• content analysis

• observational field studies

comparative studies

• examine similarities and differences between groups

  • cross-cultural research

• Challenges:

• hard to confirm equivalence of data

  • what if japan and canada use different scales to measure food security in each country

• hard to confirm Equivalence of concepts

  • some cultures will have different meanings of ideas

    • even within cultures

• equivalence of indivators

  • measurement of appitite as an indicator of nutritional status

    • hard to measure as there is no official definition

      • need to operationally define this indicator

  • use of a proxy

    • questionaires may be the best way but then they must confirm that the two palces conducting studies are measuring variables in similar ways

• selecting evidence

  • ensuring the evidence upon which comparisons are based is equivalent between groups

• Sampling

  • ensurign that the selection of analysis units is careful so that meaningful comparisons can be made

• Language

  • make sure that data is collected in different languages is still equivalent

Secondary data analysis

• analysis of published material

• analysis of data which was collected by others

strengths and limits of secondary data analysis

• cost effective

• limitations:

  • missing or incomplete information

  • can only analyze the data whcih is available

  • you do not get to determine what to analyze

Content analysis

• technique for making inferences by objectively and systematically identifying specified characteristics of messages

• Textual analysis

  • compare contrast and categorize data

• Data: written, visual, verbal

  • qualitative and quantitative

• Data analysis

  • categorizing data

    • interpretive

  • counting frequencies

Basic descisions of content analysis

• unit of analysis

• how are units selected

• what is measured and how

• how is information going to be recorded

  • coefficient of reliability - asses similarity and differences among raters

• how is informtion analyzed

Observational field studies

• field experiments

• no interaction with participants

  • how many ppl run stop signs?

Methodologies for qualitative research

• ethnography

• grounded theory

• phenomenology

Ethnography

• generatign cultural or contextual description

  • consideration of the values and practices of cultural group

  • studies behaviour of culture sharing group

  • learn from people rather than studying them

Grounded theory

• exploration of social processes

  • generation of explinations

  • consideration of experiences over time

  • generation of theory grounded in data

Phenomenology

• study of phenomena

• focuses on the meanign attached to lived experiences

• purpose is to describe the essence of lived experiences

• provides rich descriptions

• meaning of a concept or phenomenon

Reflexivity

• researchers position and background can effect

  • what they choose to study

  • how they choose to study it

  • most appropriate methods

  • interpretation or framing of results

• recognize that all knowledge is partial and situated

qualitative data collection

• participant observation

• in-depth interviews

• focus groups

Participant observation

• researcher is an active participant

• gains entery and builds rapport with the group

• data collection: field notes, memos, reflective notes

• reporting: descriptive narrative with attention to context

• exiting the field in a way that does not abandon the subjects of the study

In depth interviews

• interview guide

• emergent design

• one on one or groups

• verbatum interview transcripts for analysis

• objective rich descriptions of phenomenon from participant perspectives

Focus groups

• roots in 40s, marketing in the 70s, academics in the 80s

• Researcher is group facilitator

• 6-12 participants

• group dynamic encourages discussions of issues that would not be raised in 1-1 interviews

• collection of data from audio, video, notes etc

Strengths and limits of focus groups

• strengths

  • group dynamic

  • understanding process of dealign wiht issues

  • pilot work (needs assessment studies)

  • cost effective vs 1-1 interviews

• Limitations

  • influence of the group dynamic

  • dominant voice over quiet or shy participants

  • not generalizable

Qualitative data analysis

• code: apply conceptial label to event, action or intervention

• identification of patterns: cluster codes

• identification of themes: things that stand out, have evidence from data, confirmed by several participants

Ethnograph - data analysis software

• computer program to import text qualitative data

• helps to search note segments or interests within data

• marks them with code words so they can be retrieved for inclusion in reports or further analysis

Heierarchy of evidence based medicine

• systematic reviews are at the top

• then RCT

• Bottom is expert opinions and case/time series

SLR process

• systematic search

• critical appraisal

• synthesis of the literature

Systematic search

• identify a topic

• develop clear primary research question

  • can include intervention, population of interest, outcomes and a time reference

  • clear question should be searchable

• Establish study selection criteria

  • inclusion and exclusion (type of intervention, population etc)

• formal literature search strategy

  • online databases must be selected (done in duplicate by 2 researchers)

• document formal search strategy

  • list online databases that are searched

  • terms used

  • results retrieved

  • date searches were completed

Critical reading

• highlight the main steps of research process

• record questions

• record key study variables

• highlight new terms and significant sentences

• definitions of unfamiliar terms

Critical appraisal

• appraise the article by assessing the scientific quality of the paper

  • included its design, methods and analysis

10 criteria for appraising research studies

• review of previous research

  • how closely is literature cited in the study related to previous research

  • how recent is the review

  • are there key findings that you are aware of that ahve been omitted?

• problem and purpose

  • is it clear? tied to reviewed literature?

• hypothesis

  • clear and testable?

• method/study design

  • variables defined

  • operationalized

  • methods transparent?

• Sample population

  • who is in the sample?

  • is it clear where the sampel came from?

  • how were they selected?

• Results

  • are they consistent with study question?

• discussion

  • link to initial hypothesis?

  • mention limitations?

• Significant findings

  • are the results meaningful

• References

• overarching evaluation