Research Literature Reviews: How?

1. Select your research question(s):  a precisely stated question that guides the review. (p 21)

    

   1. Example A:  How does alcohol use affect breast cancer? (less specific)
   2. Example B:  What is the relationship between drinking two or more alcoholic beverages daily in women 65 years of age and older and breast cancer? (more specific)    
      
      Example B is a better research question due its clarity of variable (two or more alcoholic drinks per day) and target population of interest (women 65 years of age and over) (p 21).

Be specific in your vision!

Select bibliographic or articles databases, websites, and other sources (p 18-19).

Since Massasoit is a community college,  students should take advantage of the free-to-search databases on the internet

1. Google Scholar - 150 million sources - free site that mirrors Web of Science - interdisciplinary
2. Mendeley - 100 million sources - crowdsourced by authors/researchers - interdisciplinary
3. BioOne - 180 scientific journals with only free abstracts/references - life sciences.
4. Science.gov - US government clearinghouse for federal agency research - interdisciplinary
5. PubMed - US Library of Medicine - biomedical & health sciences

Choose search terms: (p.21-26)

1. Keywords - generated by the researcher from initial research question
   1. What is the relationship between drinking two or more alcoholic beverages daily in women 65 years of age and older and breast cancer? 
      1. keywords:  women 65 years of age or older, breast cancer, alcoholic beverages
      2. 50,022 results in PubMed
          
2. Thesaurus - controlled vocabulary issued by the publisher to index the database articles across disciplinary fields that may use different terms for the same concept;  generally called descriptors, subject headings, MESH terms, identifiers, etc. 
   1. What is the relationship between drinking two or more alcoholic beverages daily in women 65 years of age and older and breast cancer?
      1. PubMed's MESH terms: ("Aged"[Mesh] OR "Aged, 80 and over"[Mesh]) AND "Alcoholic Beverages"[Mesh] AND "Breast Neoplasms"[Mesh]
      2. 21 results in PubMed

Apply practical screening criteria - limits scope of literature reviewed by adding parameters for article inclusion or exclusion: (p. 53-54)

1. Publication language                                  7. Research design
2. Journal                                                        8. Sampling
3. Author                                                         9. Date of publication
4. Setting                                                        10. Date of data collection
5. Participants/subjects                                  11. Content including topics & variables
6. Program/intervention                                  12. Sources of funding

Apply methodological screening criteria to select highest quality studies possible: (p. 56-63) 

To find the highest quality articles,  reviewers should ask the following questions first (p. 56):

1. Is this study's research design internally & externally valid?
2. Are the data sources used in the study reliable & valid?
3. Are the analytical methods appropriate given the characteristics & quality of the study's data?
4. Are the results meaningful in practical & statistical terms?

Factors to keep in mind (p. 57-98):

1. Determine research study design - the way the subjects of a study are organized and measured
   1. Experimental - uses a control group vs subjects undergoing the new program or intervention - adds new knowledge - more vigorous.  Examples include:
      1. Randomized controlled trial - subjects assigned to groups randomly
      2. Parallel controls or nonrandomized controlled trials - subjects not assigned to groups randomly 
      3. Self-control or longitudinal - pretests before and posttests after intervention
      4. Historical Control - use of normative (preexisting) data compared to a group
          
   2. Observational - uses existing information to draw conclusions on past or existing conditions & activities - describes current conditions - less vigorous
      1. Cohort - provide data about changes in a specific population
      2. Case Controls - determine similarities & differences between two groups (i.e. one with a medical condition,  the other without)
      3. Cross-section - descriptive or survey data at one fixed point in time
          
2. Address bias within the study - arise from unanticipated and/or unrecognized characteristics
   1. Selection bias -  occurs when selection of experiment subjects leads to a result that is systematically different to the target population.
   2. Randomization - gold standard in reducing biases at the start of the experiment
   3. Blinding - gold standard for reducing bias during the experiment
       
3. External validity - experimental results apply to the target population - avoid the following: (p. 82-84)
   1. Reactive effects of testing - a premeasure can sensitize participants to the aims of an intervention
   2. Interactive effects of selection - an intervention and the study's participants are a unique mixture 
   3. Reactive effects of innovation - study subjects act uncharacteristically due to the artificiality of the experiment
   4. Multiple-program interference - difficulty in isolating the effects of the intervention because of the possibility that subjects are in other complementary activities or programs. 
       
4. Internal validity means the study is free from nonrandom error or bias  - below are the potential threats to study accuracy (p. 81-82)
   1. Maturation - age of subjects over course of experiment & aging's effect on the results
   2. Selection - how people are chosen & assigned to groups for the study 
   3. History - events in the world can bias study results
   4. Instrumentation - measures used to collect data are dependable
   5. Statistical regression - tendency of very high or low values to move toward the mean or average.
   6. Attrition - loss of data from subject's inability to finish study (they move, die, etc.)
       
5. Sampling - a portion or subset of a larger group called a population (p 84-98)
   1. Random sampling - every member of a target population has an equal chance of being selected - least biased
   2. Systematic sampling - selecting every 5th, 10th or 100th subject from a randomized list
   3. Stratified sampling - population is divided into subgroups (strata) and a random sample then selected from each strata
   4. Cluster sampling - naturally occurring group such as schools, clinics, cities, states, etc.
   5. Convenience samples - probability of selection is unknown

Two types of reviewed Information to ensure quality of evidence within article:

1. Methods
   1. research design
   2. sampling
   3. data collection
   4. analysis
2. Content
   1. objectives
   2. participants
   3. settings
   4. interventions
   5. results
   6. findings
   7. conclusion
3.  
4. Eligibility
   1. contains relevant information us accessible
   2. meets preset standards for methodological quality
   3. does not have any features that justify its exclusion
5. Actuality