Draft:Control groups

From Wikiversity
Jump to navigation Jump to search
The sculpture shown is part of the Boneyard Creek Sculpture series, created by Frank Gallo (control group IAS IL000018). Credit: M dorothy.

In the design of experiments, treatments are applied to experimental units in the treatment group(s).[1] In comparative experiments, members of the complementary group, the control group, receive either no treatment or a standard treatment.[2]

For the conclusions drawn from the results of an experiment to have validity, it is essential that the items or patients assigned to treatment and control groups be representative of the same population.[3] In some experiments, such as many in agriculture[4] or psychology,[5][6][7] this can be achieved by randomly assigning items from a common population to one of the treatment and control groups.[1]

The findings demonstrate a statistically systematic change from the status quo or the control group.


Def. a "number of things or persons being in some relation to one another"[8] is called a group.

The concept of a group or the concept of two or more individuals "being in some relation to one another" may be an inherited trait of all mammals, not just humans.


"After antituberculous treatment, 29 % remained positive for immune complexes, a frequency similar to that in a group of racially matched clinic controls (35%)."[9]

"Results had a normal distribution, so comparisons were made between values before and after therapy using Student's paired t tests. Unpaired t tests were used to compare clinic controls with patients with TB."[9]

As the concept of controls may have originated from the concept of controllers, the key to understanding dominant group may be solely understanding the manifestation of dominance.


  1. a "state of being [...] of prime importance; supremacy",[10]
  2. being "in a position of power, authority or ascendancy over others",[10] or
  3. a "property of [an entity] such that it suppresses the expression of [another entity]"[10]

is called dominance.

As hominins are capable of egalitarianism, the manifestation of dominance may not be inherent to hominins.

Scientific controls[edit]

"A scientific control augments integrity in experiments by isolating variables as dictated by the scientific method in order to make a conclusion about such variables. In a controlled experiment, two virtually identical experiments are conducted. In one of them, the treatment, the factor being tested is applied. In the other, the control, the factor being tested is not applied. For example, in testing a drug, it is important to carefully verify that the supposed effects of the drug are produced only by the drug itself. Doctors achieve this with a double-blind study in a clinical trial: two (statistically) identical groups of patients are compared, one of which receives the drug and one of which receives a placebo. Neither the patients nor the doctor know which group receives the real drug, which serves both to curb researchers' bias and to isolate the effects of the drug."[11]

Null groups[edit]

This is an image of a Lewis rat. Credit: Charles River Laboratories.

A null group is a group to which no treatment is applied. For example, in self-diffusion studies using a radio tracer, tracer is applied to at least two specimens, one that remains at room temperature and pressure and the second which is placed in an experimental apparatus, undergoes a temperature elevation to a specific value where it remains for a period and is subsequently lowered back down to room temperature. Thin sectioning of both samples provides an opportunity for comparison. The specimen at elevated temperature should show much greater penetration of the radio tracer into the material, while the specimen at room temperature should show little or none.

"Lewis rats underwent ligation of the left anterior descending coronary artery with direct intramyocardial injection of replication-deficient recombinant adenovirus encoding [hepatocyte growth factor] HGF ([number of individuals, n] n=10) or empty null virus as control (n=9), and animals were analyzed after six weeks. Pressure-volume conductance catheter measurements demonstrated significantly preserved contractile function in the HGF group compared with Null control animals as measured by maximum developed [left ventricle] LV pressure (79±5 versus 56±4 mm Hg, [probability] P<0.001) and maximum dP/dt (2890±326 versus 1622±159 mm Hg/sec, P<0.01). Significant preservation of LV geometry was associated with HGF treatment (LV Diameter HGF 13.1±0.54 versus Null 14.4±0.15 mm P<0.01; LV wall thickness 1.73±0.10 versus 1.28±0.07 mm P<0.01). Angiogenesis was significantly enhanced in HGF treated animals as measured by both Von Willebrand’s Factor immunohistochemical staining and a microsphere assay. [Terminal deoxynucleotidyl transferase dUTP nick end labeling] TUNEL analysis revealed a significant reduction in apoptosis in the HGF group (3.42±0.83% versus 8.36±1.16%, P<0.01), which correlated with increased [B-cell lymphoma 2] Bcl-2 and [B-cell lymphoma-extra large] Bcl-xL expression in the HGF animals."[12]

"There were ten patients with complete remissions with intensive treatment [for non-Hodgkin's lymphoma]. The most common subgroup within the complete responders were patients with cells that were nonexpressive (null) and were classified by Lukes-Collins criteria as the large, noncleaved follicular center cell variant. Currently there are seven patients remaining in complete remission, five of whom have been in continuous disease-free remission for more than two years (total survival 34+−42+ months) following cessation of all treatment."[13]

"Of those five [long term survivors], four were in the null group and four were classified as large non-cleaved. The actuarial survival curve for all null patients is characterized by a rapid initial decline and a subsequent plateau, which contains four of the long survivors."[13]

Placebo groups[edit]

Treatment applied to the group consists of a placebo so that no change is expected.

Positive groups[edit]

Treatment is applied to this group to produce an expected change, even though the change may not be a standard change.

Standard groups[edit]

"Mean nitrogen intakes (15.6 vs 9.0 gm/day) and nitrogen balances (-2.2 vs -6.6 gm/day) measured in the first 20 patients were significantly greater in the supplemented group than in the standard group (p = 0.05)."[14]

"Results: 25 patients were excluded from analysis (8 in the standard group because relevant data were not recorded; and 17 in the transumbilical group in 4 of whom relevant data were not recorded, and 13 for technical reasons)."[15]

"Those who had transumbilical cholecystectomy had significantly lower pain scores (p < 0.05) and required significantly less analgesia during the first 24 hours (p < 0.05) than those who had standard laparoscopic cholecystectomy."[15]

Randomized groups[edit]

Different experimental treatments are distributed to groups randomly to reduce systematic errors.

Blind groups[edit]

Some or all information regarding the treatment is withheld from group members.

Double-blind groups[edit]

Some or all information regarding treatments is withheld from participants and experimenters.

Wait-list groups[edit]

The group members receive treatment after the treatment group so as to serve as a control group during the study. A wait-list group is an afflicted group that receives treatment within the access to treatment window.

Planetary sciences[edit]

"Dramatically different are a control group of 13 Aten/Apollo/Amor objects selected from noncometary orbits."[16]


"Black-haired individuals were only encountered in the control group, whereas light blond and red hair color were over represented in the melanoma group. The distribution of the different eye colors (brown, gray, green, and blue) was not different between the melanoma patients".[17]


"The bone mineral content was, on average, greater for the cross-country runners than for the control group (Table 2, Figure 1)."[18]

"The cross-country runners and the controls were of similar body size, mean body height 176 and 176 cm, and mean body weight 73 and 75 kg,respectively."[18]

Control group theory[edit]

Def. "(in an experiment or clinical trial) a group of subjects closely resembling the treatment group in many demographic variables but not receiving the active medication or factor under study and thereby serving as a comparison group when treatment results are evaluated"[19] is called a control group.

Def. "any group used as a control in a statistical experiment, esp a group of patients who receive either a placebo or a standard drug during an investigation of the effects of another drug on other patients"[20] is called a control group.

Def. "in a test or trial, a group that does not receive the new treatment being studied and which is compared to the group that does receive the new treatment"[21] is called a control group.

Def. "[a] group used as a standard of comparison in a control experiment"[22] is called a control group.

Def. "(in an experiment) [t]he group of test subjects left untreated or unexposed to some procedure and then compared with treated subjects in order to validate the results of the test"[23] is called a control group.

Dominant group does not occur in any of the above dictionaries.

Dominant group[edit]

Control group is a synonym for dominant group.

A dominant group may prefer that subordinate groups, slaves (or representatives of same) be used as members of the treatment group to test a possible new treatment so that the lives or wants of the dominant group are not in jeopardy, especially where failure is a significant possibility.

Examples from primary sources are to be used to prove or disprove each hypothesis. These can be collected per subject or in general. As control group is a synonym for dominant group, testing control groups versus each hypothesis may benefit the overall effort to understand dominant group.

  1. Accident hypothesis: dominant group is an accident of whatever processes are operating.
  2. Artifact hypothesis: dominant group may be an artifact of human endeavor or may have preceded humanity.
  3. Association hypothesis: dominant group is associated in some way with the original research.
  4. Bad group hypothesis: dominant group is the group that engages in discrimination, abuse, punishment, and additional criminal activity against other groups. It often has an unfair advantage and uses it to express monopolistic practices.
  5. Control group hypothesis: there is a control group that can be used to study dominant group.
  6. Entity hypothesis: dominant group is an entity within each field where a primary author of original research uses the term.
  7. Evolution hypothesis: dominant group is a product of evolutionary processes, such groups are the evolutionary process, produce evolutionary processes, or are independent of evolutionary processes.
  8. Identifier hypothesis: dominant group is an identifier used by primary source authors of original research to identify an observation in the process of analysis.
  9. Importance hypothesis: dominant group signifies original research results that usually need to be explained by theory and interpretation of experiments.
  10. Indicator hypothesis: dominant group may be an indicator of something as yet not understood by the primary author of original research.
  11. Influence hypothesis: dominant group is included in a primary source article containing original research to indicate influence or an influential phenomenon.
  12. Interest hypothesis: dominant group is a theoretical entity used by scholarly authors of primary sources for phenomena of interest.
  13. Metadefinition hypothesis: all uses of dominant group by all primary source authors of original research are included in the metadefinition for dominant group.
  14. Null hypothesis: there is no significant or special meaning of dominant group in any sentence or figure caption in any refereed journal article.
  15. Object hypothesis: dominant group is an object within each field where a primary author of original research uses the term.
  16. Obvious hypothesis: the only meaning of dominant group is the one found in Mosby's Medical Dictionary.
  17. Original research hypothesis: dominant group is included in a primary source article by the author to indicate that the article contains original research.
  18. Primordial hypothesis: dominant group is a primordial concept inherent to humans such that every language or other form of communication no matter how old or whether extinct, on the verge of extinction, or not, has at least a synonym for dominant group.
  19. Purpose hypothesis: dominant group is written into articles by authors for a purpose.
  20. Regional hypothesis: dominant group, when it occurs, is only a manifestation of the limitations within a region. Variation of those limitations may result in the loss of a dominant group with the eventual appearance of a new one or none at all.
  21. Source hypothesis: dominant group is a source within each field where a primary author of original research uses the term.
  22. Term hypothesis: dominant group is a significant term that may require a 'rigorous definition' or application and verification of an empirical definition.


"Invasive mucinous carcinoma of the ovary (mucinous epithelial ovarian cancer [mEOC][a separate entity requiring specific treatment]) is a histologic subgroup of epithelial ovarian cancer (EOC)."[24]

"The study group consisted of all patients with advanced (International Federation of Gynecology and Obstetrics stage III or IV) mEOC treated with a first-line platinum-based therapy on the Gynecology Unit of the Royal Marsden Hospital (RMH) in London, United Kingdom, between September 1992 and January 2001. For each case, two controls with advanced non-mEOC undergoing first-line platinum-based therapy were randomly chosen from the unit's database. Controls were matched for date of diagnosis and stage of disease, in this order of priority."[24]

"The median age was 58 years (range, 32 to 76 years) for the study group and 60 years (range, 26 to 76 years) for the control group; this difference was not statistically significant. Nineteen percent of patients with mEOC and 13% of patients in the control group had stage IV disease; this difference was not statistically significant. The percentage of patients who were optimally debulked by surgery (< 2 cm of macroscopic disease) was similar in both the cases and the control group."[24]

Gamma rays[edit]

"Because of the larger number of pulsars in this study it is possible to do a thorough population study of pulsars in this energy range [100 MeV to 10 GeV] by dividing the pulsars into a candidate group (high Ė/d2) and a control group (low Ė/d2)."[25] Bold added. Specifically, "[s]ince such a large number of pulsars are involved in this study, it is possible to divide them into two groups: one containing the least likely candidates, which we will call the "control" group and the group of best candidates (called the "candidate" group). The division is made by means of the gamma-ray pulsar model-indpendent parameter Ė33/d2kpc, where Ė33 is the rotational energy loss rate of the star in units of 1033 ergs s-1 and dkpc is the distance to the source in kiloparsecs."[25]


"The cognitive theory of the young Steiner is at one and the same time an ontology and a cosmogony—a regression to the pre-modern naive movement of universal realism. Its aim is to show man his task and position in the universe through a process of self-reflection and to ensure that through the thought process [...] man is able to achieve something which he once owed to a belief in revelation, namely the satisfaction of the mind.6"[26]

"Steiner’s cosmogony takes the basic form of the gnostic myth: man must lose his worldliness and slavish dependence on material things so that the soul and the world can rise up to self-redemption and fuse once again with the divine spiritual origins which both bear within them. Modern man lives on the fourth planetary phase of development of the earth that entails an experience of individuation and the respiritualization of the individual."[26]

"An earlier quantitative survey of former German pupils of the Rudolf Steiner schools (born in the year 1940/41) revealed significant differences between this group and a control group in the following areas: higher geographical and social mobility; more pronounced leisure activities in the areas of reading, interest in art, practice of a musical instrument and ability in craftwork; and an interest in further training.21"[26]


Take identical growing plants and give fertilizer to half of them. If there are differences between the fertilized plant group and the unfertilized control group, these differences may be due to the fertilizer. Credit: Forest & Kim Starr.

Test plant fertilizer by giving it to only half the plants in a garden: the plants that receive no fertilizer are the control group, because they establish the baseline level of growth that the fertilizer-treated plants will be compared against. Without a control group, the experiment cannot determine whether the fertilizer-treated plants grow more than they would have if untreated.


"This paper constitutes an argument for the use of quasi-experimental control group methods as a measurement technique to study economic and spatial structural change. The essence of such methods is the careful identification of a control group--a set of places whose economic development enables measurement of what would have happened in the place under study without the phenomenon or policy being studied."[27]

"A group of places where something did not occur (no treatment) can be selected nonrandomly to be a control group for a place or places where something did occur (treatment)."[27]


"[T]he northern outcrop area where stratigraphic control is very good, provided the control group to determine whether each of the four beds carries a unique chemical fingerprint".[28]

"Control groups provide a qualitative means of examining the completeness of the record and thus the validity of crocodilian patterns."[29]

"In this study, noncrocodilian vertebrates provide a broad-based control group; thus, the absence of crocodilians at a vertebrate locality is considered to reflect environmental factors (including climate) and not just to be an artifact of collection or taphonomic biases."[29]

"Cases [are] correctly classified within the control group of felsic igneous (Group 1), intermediate igneous (Group 2), and marie igneous (Group 3)."[30]


"[G]ains for people of color come when such a gain would advance the interests of society's control group and do not impose too great a cost on other whites in a position to resist them."[31]


"Matched sampling is a method for selecting units from a large reservoir of potential controls to produce a control group of modest size that is similar to a treated group with respect to the distribution of observed covariates."[32]

"Many educators believe that [computer-based education] CBE is the panacea for education because of the education theories used in the development of the computer. The demands of World War II helped to accelerate the development of the computer. ... In most of the meta-analyses in this study, the h statistic was used to measure the effects of CBE on instructional time and was calculated by dividing the instructional time required of the experimental group by the instruction time required of the control group."[33]


In some medical studies, where it may be unethical not to treat patients who present with symptoms, controls may be given a standard treatment, rather than no treatment at all.[2] Another alternative is to select controls from a wider population, provided that this population is well-defined and that those presenting with symptoms at the clinic are representative of those in the wider population.[3]


"Relative to a control group of students studying literature, autism occurred significantly more often in families of students in the fields of physics, engineering, and mathematics."[34]


A "clinic control group of 20 (12 M, eight F) TB contacts or recent immigrants to Brent. (Mean age 34 years (range 18-70); Indian 16 (80%), Caucasian three (15%), and Negro one (5%).)"[9]

"Patients with intra- and extra-thoracic disease both had similar amounts of complexes. Similarly, age, sex, and race had no influence on their presence, though they were present in the sera of 7/20 (35 %) of the clinic control group, comprising largely recent immigrants to this country [England]."[9]


"A laboratory control group of sera from 80 laboratory and medical personnel was used to define the normal limits for the presence of immune complexes."[9]

"The amounts of Clq, IgG, and IgM in the precipitate were measured directly by single radial immunodiffusion (SRID), and the presence of IgA was determined by double diffusion. The values were compared initialy with those obtained in this test from the laboratory control group. To study the effect of serum immunoglobulin levels on the amount of precipitated immunoglobulin 217 sera from patients with a variety of diseases were studied."[9]

"After therapy, levels of Clq, IgG, and IgM complexes were similar to those in both the laboratory and clinic control groups."[9]

"There was a significantly higher frequency of circulating immune complexes in the control group drawn from the Chest Clinic patients. They were matched for race with the patients studied and in so far as racial origin was concerned they differed significantly from the subjects who comprised the laboratory controls."[9]

"The levels in the treated and clinically cured patients came down not to the normal level for the laboratory control group but to the normal level for our local population."[9]


  1. A control group can be used to study control groups.

See also[edit]


  1. 1.0 1.1 Klaus Hinkelmann, Oscar Kempthorne (2008). Design and Analysis of Experiments, Volume I: Introduction to Experimental Design (2nd ed.). Wiley. ISBN 978-0-471-72756-9.
  2. 2.0 2.1 R. A. Bailey (2008). Design of comparative experiments. Cambridge University Press. ISBN 978-0-521-68357-9. MR 2422352.
  3. 3.0 3.1 Everitt, B.S. (2002) The Cambridge Dictionary of Statistics, CUP. ISBN 0-521-81099-X (entry for control group)
  4. Jerzy Neyman (1923). Dabrowska, Dorota M.; Speed, Terence P.. eds. "On the application of probability theory to agricultural experiments: Essay on principles (Section 9)". Statistical Science 5 (4): 465–472. 
  5. Ian Hacking (September 1988). "Telepathy: Origins of Randomization in Experimental Design". Isis (A Special Issue on Artifact and Experiment) 79 (3): 427–451. 
  6. Stephen M. Stigler (November 1992). "A Historical View of Statistical Concepts in Psychology and Educational Research". American Journal of Education 101 (1): 60–70. doi:10.1086/444032. 
  7. Trudy Dehue (December 1997). "Deception, Efficiency, and Random Groups: Psychology and the Gradual Origination of the Random Group Design". Isis 88 (4): 653–673. doi:10.1086/383850. PMID 9519574. 
  8. group. San Francisco, California: Wikimedia Foundation, Inc. 17 February 2015. Retrieved 2015-03-21.
  9. 9.0 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 NMcI Johnson, MW McNicol, EJ Burton-Kee, JF Mowbray (1981). "Circulating immune complexes in tuberculosis". Thorax 36 (8): 610-7. http://thorax.bmj.com/content/36/8/610.full.pdf. Retrieved 2014-05-27. 
  10. 10.0 10.1 10.2 dominance,. San Francisco, California: Wikimedia Foundation, Inc. 24 August 2014. Retrieved 2015-03-21.
  11. C Michael Gibson (14 October 2007). Scientific control. WikiDoc. Retrieved 2018-06-04.
  12. Vasant Jayasankar, Y. Joseph Woo, Lawrence T. Bish, Timothy J. Pirolli, Subhasis Chatterjee, Mark F. Berry, Jeffrey Burdick, Timothy J. Gardner, H. Lee Sweeney (September 9, 2003). "Cardiac Transplantation and Surgery for Congestive Heart Failure Gene Transfer of Hepatocyte Growth Factor Attenuates Postinfarction Heart Failure". Circulation 108 (10 Supplement 1): II-230-6. doi:10.1161/​01.cir.0000087444.53354.66. PMID 12970238. http://circ.ahajournals.org/content/108/10_suppl_1/II-230.full. Retrieved 2014-05-26. 
  13. 13.0 13.1 Richard A. Rudders, Ernest T. Ahl Jr, and Ronald A. Delellis (March 15, 1981). "Surface marker and histopathologic correlation with long‐term survival in advanced large‐cell non‐Hodgkin's lymphoma". Cancer 47 (6): 1329-35. http://onlinelibrary.wiley.com/doi/10.1002/1097-0142(19810315)47:6%3C1329::AID-CNCR2820470616%3E3.0.CO;2-X/abstract. Retrieved 2014-05-27. 
  14. JM Daly, MD Lieberman, J Goldfine, J Shou, F Weintraub, EF Rosato, and P Lavin (1992). "Enteral nutrition with supplemental arginine, RNA, and omega-3 fatty acids in patients after operation: immunologic, metabolic, and clinical outcome". Surgery 112 (1): 56-67. PMID 1377838. http://europepmc.org/abstract/MED/1377838. Retrieved 2014-05-27. 
  15. 15.0 15.1 Fabrizio Bresadola , Alberto Pasqualucci , Annibale Donini , Paolo Chiarandini , Gabriele Anania , Giovanni Terrosu , Marco A. Sistu and Alberto Pasetto (1999). "Elective transumbilical compared with standard laparoscopic cholecystectomy". The European Journal of Surgery 165 (1): 29-34. doi:10.3109/110241599750007478. http://informahealthcare.com/doi/abs/10.3109/110241599750007478. Retrieved 2014-05-27. 
  16. William K. Hartmann, David J. Tholen, Dale P. Cruikshank (January 1987). "The relationship of active comets,“extinct” comets, and dark asteroids". Icarus 69 (1): 33-50. doi:10.1016/0019-1035(87)90005-4. http://www.sciencedirect.com/science/article/pii/0019103587900054. Retrieved 2014-05-27. 
  17. Cornelis Kennedy, Jeanet ter Huurne, Marjo Berkhout, Nelleke Gruis, Maarten Bastiaens, W Bergman, R Willemze and Jan Nico Bouwes Bavinck (2001). "Melanocortin 1 Receptor (MC1R) Gene Variants are Associated with an Increased Risk for Cutaneous Melanoma Which is Largely Independent of Skin Type and Hair Color". Journal of Investigative Dermatology 117 (2): 294-300. http://www.nature.com/jid/journal/v117/n2/abs/5601156a.html. Retrieved 2014-05-27. 
  18. 18.0 18.1 Nils Dalén and Karl Erik Olsson (1974). "Bone mineral content and physical activity". Acta Orthopaedica Scandica 45: 170-4. doi:10.3109/17453677408989136. http://informahealthcare.com/doi/pdf/10.3109/17453677408989136. Retrieved 2014-05-27. 
  19. Random House (2012). control group noun, based on the Random House Dictionary. Random House. Retrieved 2012-04-06.
  20. Harper Collins (2012). control group, In: Collins English Dictionary - Complete & Unabridged 10th Edition. HarperCollins Publishers. Retrieved 2012-04-06.
  21. Twenty-first Century (2012). control group, In: Dictionary.com's 21st Century Lexicon. Retrieved 2012-04-06.
  22. Stedman (2012). control group, In: The American Heritage® Stedman's Medical Dictionary. Houghton Mifflin Company. Retrieved 2012-04-06.
  23. control group. San Francisco, California: Wikimedia Foundation, Inc. October 13, 2012. Retrieved 2013-03-16.
  24. 24.0 24.1 24.2 Viviane Hess, Roger A'Hern, Nazar Nasiri, D. Michael King, Peter R. Blake, Desmond P.J. Barton, John H. Shepherd, T. Ind, J. Bridges, K. Harrington, Stanley B. Kaye and Martin E. Gore (March 15, 2004). "Mucinous Epithelial Ovarian Cancer: A Separate Entity Requiring Specific Treatment". Journal of Clinical Oncology 22 (6): 1040-4. doi:10.1200/JCO.2004.08.078. http://jco.ascopubs.org/content/22/6/1040.full. Retrieved 2014-05-28. 
  25. 25.0 25.1 H. I. Nel, Z. Arzoumanian, M. Bailes, K. T. S. Brazier, N. D'Amico, J. A. Esposito, C. E. Fichtel, J. M. Fierro, S. D. Hunter, S. Johnston, G. Kanbach, V. M. Kaspi, D. A. Kniffen, Y. C. Lin, A. G. Lyne, R. N. Manchester, J. R. Mattox, H. A. Mayer-Hasselwander, M. Merck, P. F. Michelson, D. J. Nice, P. L. Nolan, P. V. Ramanamurthy, J. H. Taylor, D. J. Thompson, and C. Westbrook (July 10, 1996). "EGRET High-Energy Gamma-Ray Pulsar Studies. III. A Survey". The Astrophysical Journal 465 (07): 898-906. doi:10.1086/177473. http://adsabs.harvard.edu//abs/1996ApJ...465..898N. Retrieved 2013-03-16. 
  26. 26.0 26.1 26.2 Heiner Ullrich (1994). "Rudolf Steiner". Prospects: the quarterly review of comparative education 24 (3/4): 555-72. http://www.ibe.unesco.org/publications/ThinkersPdf/steinere.pdf. Retrieved 2013-12-18. 
  27. 27.0 27.1 AM Isserman, JD Merrifield (January 1987). "Quasi-experimental control group methods for regional analysis: An application to an energy boomtown and growth pole theory". Economic geography 63 (1): 3-19. http://www.rri.wvu.edu/pdffiles/wp8610.pdf. Retrieved 2013-03-17. 
  28. Dennis R. Kolata, Joyce K. Frost and Warren D. Huff (March 1987). "Chemical correlation of K-bentonite beds in the Middle Ordovician Decorah Subgroup, upper Mississippi Valley". Geology 15 (3): 208-11. doi:10.1130/​0091-7613(1987)​15<208:CCOKBI>​2.0.CO;2. http://geology.geoscienceworld.org/content/15/3/208.short. Retrieved 2013-03-17. 
  29. 29.0 29.1 Paul J. Markwick (March 1994). ""Equability," continentality, and Tertiary "climate": The crocodilian perspective". Geology 22 (7): 613-6. doi:10.1130/0091-7613(1994)​022<0613:ECATCT>​2.3.CO;2. http://geology.gsapubs.org/content/22/7/613.short. Retrieved 2013-03-17. 
  30. Jeffry D. Grigsby (March 1992). "Chemical Fingerprinting in Detrital Ilmenite: A Viable Alternative in Provenance Research?". Journal of Sedimentary Petrology 62 (2): 331-7. http://archives.datapages.com/data/sepm/journals/v59-62/data/062/062002/0331.htm. Retrieved 2013-03-17. 
  31. Richard Delgado (1991). "Recasting the American Race Problem". California Law Review 79: 1389. http://heinonlinebackup.com/hol-cgi-bin/get_pdf.cgi?handle=hein.journals/calr79&section=49. Retrieved 2012-05-31. 
  32. PR Rosenbaum (February 1985). "Constructing a control group using multivariate matched sampling methods that incorporate the propensity score". American Statistician 39 (1): 33-8. doi:10.2307/2683903. http://www.stat.duke.edu/~banks/130-lectures.dir/readbyNov22.pdf. Retrieved 2012-05-31. 
  33. Janis Lowe (Winter 2001). "Computer-based education: Is it a panacea?". Journal of Research on Technology in Education 34 (2): 163-71. http://search.ebscohost.com/login.aspx?direct=true&profile=ehost&scope=site&authtype=crawler&jrnl=15391523&AN=16830813&h=P9CzTOw07Hl9rVH2a07sDaB95%2BvhqELDZJHAh2C1dPdZF3rBh%2FEt%2FWliOeBfvbYboh%2BrOFfjigVD46OLPkcpYQ%3D%3D&crl=c. Retrieved 2014-05-28. 
  34. Simon Baron-Cohen, Patrick Bolton, Sally Wheelwright, Victoria Scahill, Liz Short, Genevieve Mead, and Alex Smith (1998). "Autism occurs more often in families of physicists, engineers, and mathematicians". Autism 2: 296-301. http://citeseerx.ist.psu.edu/viewdoc/download?doi= Retrieved 2013-03-17. 

External links[edit]

{{Humanities resources}}