Report Jan 24
-
Upload
faye-binuya -
Category
Documents
-
view
217 -
download
0
Transcript of Report Jan 24
-
8/3/2019 Report Jan 24
1/98
Chapter 3
Methods and Procedures
-
8/3/2019 Report Jan 24
2/98
SHIELA S. SAGMITDMD 3AA
-
8/3/2019 Report Jan 24
3/98
OBSERVATIONAL METHODy Aphenomena is being observed and recorded
y Studies which could be defined as observational
research including case studies, ethnographic studies,
ethological studies, etc.
-
8/3/2019 Report Jan 24
4/98
CORRELATIONAL METHODy Examines the covariation of two or more variables.
Example:
The early research on cigarette smoking examine
the covariation of cigarette smoking and a variety of
lung disease.
-
8/3/2019 Report Jan 24
5/98
CORRELATIONAL METHODyCan be accomplished by a variety of techniques
which include the collection of emprical data.
yNothing is manipulated by the experimenter.
y
Not casual research.y Exploratory or beginning research.
-
8/3/2019 Report Jan 24
6/98
TRUE EXPERIMENTS /
EXPERIMENTALWhat is the cause?
y At least one variable is manipulates and its effects aremeasured
y Subjects randomly assigned to experimental treatmentand control group
y Who are treated the same except for the treatmentvariable determined cause and effect
y (When intact groups are used its called quasi experimental)
-
8/3/2019 Report Jan 24
7/98
TRUE EXPERIMENTS
y Laboratory study
y Experiment conducted where an effort is made to
impose control over all other variables except the one
under study.
-
8/3/2019 Report Jan 24
8/98
TO UNDERSTAND THE
NATURE OF THEEXPERIMENT, WE MUST
FIRST DEFINE FEWTERMS
-
8/3/2019 Report Jan 24
9/98
1.
Experimental or treatmentgroup
The group that receives the experimentaltreatment, manipulation, or is different
from the control group on the variable under
study.
-
8/3/2019 Report Jan 24
10/98
2. Control group
Agroup that is used to produce comparisons. The
treatment of interest is deliberately withheld or
manipulated to provide a baseline performance
with which to compare the experimental or
treatment group's performance.
-
8/3/2019 Report Jan 24
11/98
3. Independent variable
The variable that the experimenter manipulates in
a study. It can be any aspect of the environment
that is empirically investigated for the purpose of
examining its influence on the dependent variable.
-
8/3/2019 Report Jan 24
12/98
4. Dependent variableThe variable that is measured in a study.
The experimenter does not control this
variable.
-
8/3/2019 Report Jan 24
13/98
5. Random assignmentEach subject has an equal probability of
being selected for either the treatment or
control group.
-
8/3/2019 Report Jan 24
14/98
6. Double blind
Neither the subject nor the experimenterknows whether the subject is in the
treatment of the control condition
-
8/3/2019 Report Jan 24
15/98
QUASI-EXPERIMENTSyvery similar to true experiments but use naturally
formed or pre-existing groups.
Naturally formed groups - the variable under study is a
subject variable.Pre-existing groups - the variable that is manipulated
between the two groups is an independent variable
-
8/3/2019 Report Jan 24
16/98
QUANTITATIVE OR DESCRIPTIVE
What is the current situation?
y Numerical data gathered through tests, surveys, observations,
interviews.
y Variables are not manipulated but are measured as they occur
y Subgroups may be compared on some measure
y Two or more variables of a group may be correlated
y Does not attempt to identify cause of differences or relationships
-
8/3/2019 Report Jan 24
17/98
EX POST FACTO/CASUAL COMPARATIVEWhat is the possible cause?
y Identifies an effect that has already occurred and attempts
to infer causey Atreatment variable (alleged cause) is identified (but not
manipulated) and effects are measured
y Groups exposed to the treatment variable are compared to
groups who are noty Identification of cause can be called into question because
groups were not randomly assigned and other extraneousvariables were not controlled
-
8/3/2019 Report Jan 24
18/98
QUALITATIVE OR HISTORICALWhat was the situation?
y
Description of past events, problems, issues, facts, datagathered from written or oral descriptions of past events,
artifatcs, etc.
y
Describes what was in an attempt to reconstruct the pasty Involves much interpretation of events and its influence on
the present
-
8/3/2019 Report Jan 24
19/98
ETHNOGRAPHICWhat is the current situation?
y I
ndepth analytical description of educational systems,process and phenomena within a specific context
based on detailed observations and interviews
y Detailed examination of single group, individual,
situation, or site is called a case study
-
8/3/2019 Report Jan 24
20/98
Respondents/ Subjects
-
8/3/2019 Report Jan 24
21/98
Sample size
The number of units (persons, animals, patients, specifiedcircumstances, etc.) in a population to be studied. Thesample size should be big enough to have a high likelihoodof detecting a true difference between two groups.
In statistics and survey methodology, sampling is concerned
with the selection of a subset of individuals from within apopulation to estimate characteristics of the wholepopulation.
Researchers rarely survey the entire population because thecost of a census is too high. The three main advantages ofsampling are that the cost is lower, data collection is faster,
and since the data set is smaller it is possible to ensurehomogeneity and to improve the accuracy and quality ofthe data.
-
8/3/2019 Report Jan 24
22/98
Sample sizeEach observation measures one or more properties
(such as weight, location, color) of observable bodiesdistinguished as independent objects or individuals. In
survey sampling, weights can be applied to the data toadjust for the sample design, particularly stratifiedsampling (blocking). Results from probability theory andstatistical theory are employed to guide practice. Inbusiness and medical research, sampling is widely used forgathering information about a population.
-
8/3/2019 Report Jan 24
23/98
Sample sizeIn order to have confidence that your survey results are
representative, it is critically important that you have a largenumber of randomly-selected participants in each group yousurvey. So what exactly is "a large number?" For a 95%confidence level (which means that there is only a 5% chanceof your sample results differing from the true populationaverage), a good estimate of the margin of error (orconfidence interval) is given by 1/N, whereNis the numberof participants or sample size (Niles, 2006).
-
8/3/2019 Report Jan 24
24/98
-
8/3/2019 Report Jan 24
25/98
-
8/3/2019 Report Jan 24
26/98
Respondents of the researchy are people who agree to take part in a research project
such as a survey. For example, if you complete a
questionnaire about your working life, and then sendit back to a student or academic who uses it to gaininformation about working life in your particularsector.
-
8/3/2019 Report Jan 24
27/98
Ex
plain how the subjects were recruited, then list means andstandard deviations of the relevant demographics (age,weight, height) plus any other relevant characteristics(recent best performances, recent training). Show ranges ofcharacteristics only if there are unusually distant outliers inthe sample. If possible, report recent best competitive
performances of athletes as a percent of the world record,to make it clear what caliber of athlete the outcome of yourstudy can be generalized to.
Show all the above characteristics for any major subgroups of
subjects (e.g., males and females, non-athletes andathletes). Include the number of subjects in each subgroup.
-
8/3/2019 Report Jan 24
28/98
-
8/3/2019 Report Jan 24
29/98
Sampling methodsare classified as eitherprobability or non-probability. In
probability samples, each member of the population has aknown non-zero probability of being selected. Probabilitymethods include random sampling, systematic sampling, andstratified sampling. In non-probability sampling, membersare selected from the population in some nonrandommanner. These include convenience sampling, judgmentsampling, quota sampling, and snowball sampling.
-
8/3/2019 Report Jan 24
30/98
Random samplingIs the purest form of probability sampling. Each member
of the population has an equal and known chance of
being selected.W
hen there are very large populations,it is often difficult or impossible to identify everymember of the population, so the pool of availablesubjects becomes biased
-
8/3/2019 Report Jan 24
31/98
1. Get a list or sampling frame
a. This is the hard part! It must not systematically
ex
clude anyone.b. Remember the famous sampling mistake?
2. Generate random numbers
3. Select one person per random number
-
8/3/2019 Report Jan 24
32/98
Systematic samplingIs often used instead of random sampling. It is also called an
Nth name selection technique. After the required sample sizehas been calculated, every Nth record is selected from a list of
population members. As long as the list does not contain anyhidden order, this sampling method is as good as the randomsampling method. Its only advantage over the randomsampling technique is simplicity. Systematic sampling isfrequently used to select a specified number of records from acomputer file.
-
8/3/2019 Report Jan 24
33/98
1. Select a random number, which will be known as k2. Get a list of people, or observe a flow of people (e.g.,
pedestrians on a corner)
3. Select everykth persona. Careful that there is no systematic rhythm to the flow or list
of people.b. If every 4th person on the list is, say, rich or senior or
some other consistent pattern, avoid this method
-
8/3/2019 Report Jan 24
34/98
Stratified samplingIs commonly used probability method that is superior to random
sampling because it reduces sampling error. Astratum is a subset of thepopulation that share at least one common characteristic. Examples ofstratums might be males and females, or managers and non-managers.
The researcher first identifies the relevant stratums and their actualrepresentation in the population. Random sampling is then used toselect a sufficient number of subjects from each stratum. "Sufficient"refers to a sample size large enough for us to be reasonably confidentthat the stratum represents the population. Stratified sampling is oftenused when one or more of the stratums in the population have a lowincidence relative to the other stratums.
-
8/3/2019 Report Jan 24
35/98
1. Separate your population into groups or strata
2. Do either a simple random sample or systematic
random sample from therea. Note you must know easily what the strata are before
attempting this
b. If your sampling frame is sorted by, say, school district,
then youre able to use this method
-
8/3/2019 Report Jan 24
36/98
Multi stage sampling1. Get a list of clusters, e.g., branches of a company
2. Randomly sample clusters from that list
3. Have a list of, say, 10 branches4. Randomly sample people within those branches
a. This method is complex and expensive!
-
8/3/2019 Report Jan 24
37/98
Convenience samplingis used in exploratory research where the researcher is
interested in getting an inexpensive approximation ofthe truth. As the name implies, the sample is selectedbecause they are convenient. This nonprobabilitymethod is often used during preliminary researchefforts to get a gross estimate of the results, withoutincurring the cost or time required to select a random
sample.
-
8/3/2019 Report Jan 24
38/98
Judgment samplingis a common nonprobability method. The researcher
selects the sample based on judgment. This is usuallyand extension of convenience sampling. For example, a
researcher may decide to draw the entire sample fromone "representative" city, even though the populationincludes all cities. When using this method, theresearcher must be confident that the chosen sampleis truly representative of the entire population.
-
8/3/2019 Report Jan 24
39/98
Quota samplingis the nonprobability equivalent of stratified sampling.Like stratified sampling, the researcher first identifiesthe stratums and their proportions as they are
represented in the population. Then convenience orjudgment sampling is used to select the requirednumber of subjects from each stratum. This differsfrom stratified sampling, where the stratums are filledby random sampling.
-
8/3/2019 Report Jan 24
40/98
1. Determine what the population looks like in termsof specific qualities.
2.C
reate quotas based on those qualities.3. Select people for each quota.
-
8/3/2019 Report Jan 24
41/98
Snowball samplingis a special nonprobability method used when the
desired sample characteristic is rare. It may be extremelydifficult or cost prohibitive to locate respondents in thesesituations. Snowball sampling relies on referrals from initialsubjects to generate additional subjects. While this techniquecan dramatically lower search costs, it comes at the expense ofintroducing bias because the technique itself reduces thelikelihood that the sample will represent a good cross sectionfrom the population.
-
8/3/2019 Report Jan 24
42/98
1. Find a few people that are relevant to your topic.
2. Ask them to refer you to more of them.
-
8/3/2019 Report Jan 24
43/98
ExamplesExample 1
Suppose you were interested in investigating the link betweenthe family of origin and income and your particular interest is incomparing incomes of Hispanic and Non-Hispanic respondents. For
statistical reasons, you decide that you need at least 1,000 non-Hispanics and 1,000 Hispanics. Hispanics comprise around 6 or 7% ofthe population. If you take a simple random sample of all races thatwould be large enough to get you 1,000 Hispanics, the sample sizewould be near 15,000, which would be far more expensive than amethod that yields a sample of 2,000. One strategy that would be morecost-effective would be to split the population into Hispanics and non-
Hispanics, then take a simple random sample within each portion(Hispanic and non-Hispanic).
-
8/3/2019 Report Jan 24
44/98
Example 2Let's suppose your sampling frame is a large city's
telephone book that has 2,000,000 entries. To take a SRS, youneed to associate each entry with a number and choose n=200 numbers from N= 2,000,000. This could be quite anordeal. Instead, you decide to take a random start between 1and N/n= 20,000 and then take every 20,000th name, etc.This is an example of systematic sampling.
-
8/3/2019 Report Jan 24
45/98
Example 3Suppose you wanted to study dance club and bar employees in NYCwith a
sample of n = 600. Yet there is no list of these employees from which to
draw a simple random sample. Suppose you obtained a list of allbars/clubs in NYC. One way to get this would be to randomly sample300 bars and then randomly sample 2 employees within each bars/club.This is an example of cluster sampling. Here the unit of analysis(employee) is different from the primary sampling unit (the bar/club).
In each of these three examples, a probability sample is drawn, yet none isan example of simple random sampling.
Although simple random sampling is the ideal for social science and mostof the statistics used are based on assumptions of SRS, in practice, SRSare rarely seen. It can be terribly inefficient, and particularly difficultwhen large samples are needed. Other probability methods are morecommon. Yet SRS is essential, both as a method and as an easy-to-understand method of selecting a sample.
To recap, though, that simple random sampling is a sampling procedurein which every element of the population has the same chance of beingselected and every element in the sample is selected by chance.
-
8/3/2019 Report Jan 24
46/98
Research
Instruments
-
8/3/2019 Report Jan 24
47/98
Research Instruments
a. questionnaire
b. interviewc. observation
d. records/documents
-
8/3/2019 Report Jan 24
48/98
Questionnaires
Questionnaires commonlyrequiresubjectstorespondtoastimulus. However,theyhavetheir
uses,especiallyasameansofcollectinginformationfromawidersamplethan canbereachedbypersonalinterview. Thoughtheinformationisnecessarilymorelimited,it canstillbeveryuseful.
-
8/3/2019 Report Jan 24
49/98
Aquestionnaireisanintegralpartofasurveymethodology. therearegenerallytwotypesofquestionnaireitems:
y open-endedand
y restrictedor close-endeditems
-
8/3/2019 Report Jan 24
50/98
y Ifthepurposeofyoursurveyistogetanadequatepictureofhowtherespondentfeelsaboutthetopic,whatitmeanstohimandthebackgroundofhisanswer,theopen-endedquestionsmaybeused . Thistechniquehowever,presentsgreatdifficultiesoftimeandexpenseintabulatingandsummarizing.
y
Ontheotherhand,usingarestrictedor close-endedquestionissimplyprovidingasetofcategoriesfortherespondentto check forfrequency. Thisformistime-saving,exercisesadirectiveinfluenceinsecuringresponses,andgreatlyfacilitatestheprocessoftabulating
andsummarizing. However,theinformationobtainedmaynotbeasrichastheinformationfromanopen-endedquestion.
-
8/3/2019 Report Jan 24
51/98
y Someusepartiallyopenendeditem,whereanother categoryisprovidedtogivetherespondentanopportunitytospecifyanitem.
Example:
Whatdoyoudoduringyourleisuretime? Encirclethenumberofallthatapply.
1 Watch TV, DVD
2 Readbooksandmagazines
3 Playoutdoorsports
4 Surftheinternet
5 Visitfriends
6 Others (specify) ____________
-
8/3/2019 Report Jan 24
52/98
y Somerestrictedquestionsusearatingscaleratherthanresponsealternatives.
Ex
ample:1_____ 2_____ 3_____ 4_____ 5 _____ Very often Often Sometimes Rarely Never
-
8/3/2019 Report Jan 24
53/98
-
8/3/2019 Report Jan 24
54/98
The following are tips designed to
help you construct a goodquestionnaire:
y Keepthewordingofyouritemssimple.
y Strivetomakeyourquestionsprecise.
y Avoidbiasedwording.
y Avoiddoublequestions.
y Avoidquestionsthatincludeanegative
-
8/3/2019 Report Jan 24
55/98
Administering a QuestionnaireAdministrationofquestionnaireinvolves
threeimportantsteps:
y Pre-testing
y Distributing/mailingquestionnaires
y Makingfollow-ups
-
8/3/2019 Report Jan 24
56/98
yItisdesirabletotryoutafew copiesofthequestionnaireandtoexaminethereturnsbeforetheinstrumentisusedonalargescale. Thispre-testwillprobablyleadtorevisionofcertainitems.
Whenthevalidityofthequestionnairehasbeentriedandtested,it canbedistributedtoitsrespondents. Toensureahighpercentageofreturns,theresearchershouldmakefollow-upprocedures.
-
8/3/2019 Report Jan 24
57/98
InterviewyWhen an interview is used as a research technique, theprocess of the dialogue between the experimenter andthe subject is a part of the experimental conditions.The dialogue can be extensively structured, or it can be
open-ended. Though the latter situation may providemore information about the subject's thoughtprocesses, there is a greater possibility of certainpitfalls, such as the experimenter missing the subject'sthoughts by anticipating them and taking too strong alead in the discussion. Another problem is introducedby the need to go beyond the point where theindividual does not know any more.
-
8/3/2019 Report Jan 24
58/98
y If a subject admits to not knowing any more, theexperimenter must question further without insultingthe subject. Athird problem is created by the subject'sdependency on the experimenter. The experimenter
must balance a neutral questioning technique withinterpretive judgments about the subject's responses.Finally, the data collection might most advantageouslybe separated into two activities: the interview and the
analysis of the transcripts, including the analysis of theinterviewer's own part in the dialogue.
-
8/3/2019 Report Jan 24
59/98
Observationy
1) Observationaltechniquesareanimportantaspectofmanyactionresearchstudiesandofcasestudieswhetherundertakenbyparticipantsoroutsiders.
y
2) Inawayallofusarealreadywellpractisedintheartsofobservation- weallneedtoobservehumanbehaviourinourpersonalandprofessionallives,weareallfamiliarwiththe
needto cometo conclusionsbasedonourobservation,togenerateexplanationsandunderstandingsandevento comeupwithpredictions.
-
8/3/2019 Report Jan 24
60/98
The observational method in a non-experimentaldesign. The absence of an independent variable doesnot allow any cause-effect conclusions to be drawnfrom observational research. Sound evidence is
however important to the observational method.Indeed, the observational method's key featureisba standardised, planned, andsystematic approach toobjectively observe and record behaviour. This is of
course to generate all-important data upon which tobase any conclusions.
-
8/3/2019 Report Jan 24
61/98
Observations, are of five main
typesy Participantobservation is where a researcher sets up
and takes part in the observational study.
y N
on-participantobservation is where theresearcher sets up but does not take part in theobservational study. They observe participants at adistance.
y
Structuredobservation is the planned watching andrecording of behaviours as they occur withina controlledenvironment. Used particularly withinfants and young children.
-
8/3/2019 Report Jan 24
62/98
y Unstructuredobservation is the unplanned,informal, watching and recording of behaviours asthey occur in a natural environment.
y N
aturalistic observation is the planned watchingand recording of behaviours as they occur withina naturalenvironment. An example would benaturalistic observation of animals in their naturalhabitat.
-
8/3/2019 Report Jan 24
63/98
Each involves the planned gathering, analysis, and
interpretation of mostly empirical data on observed
behaviour. Each observation has its own features,advantages and disadvantages. Participant observation,
for example, sees the researcher set up, and take part in
the observation of behaviour under investigation. Non-
participant observation sees no involvement on the part
of the researcher, with recordings of observed
behaviours being taken from afar. If the researcher plans,
structures, and conducts their observation appropriately,
the observational method can be seen as a most valid
and reliable form of non-experimental research inpsychology mainly due to the observational method's
high ecological validity
-
8/3/2019 Report Jan 24
64/98
Observational research
(experimental)y This type of research draws a conclusion by comparing subjects
against a control group, in cases where the researcher hasno control over the experiment.
y Aresearch study comparing the risk of developing lung cancer,between smokers and non-smokers, would be a good example ofan observational study.
y With the smoking example, a scientist cannot give cigarettes tonon-smokers for 20 years and compare them with a controlgroup. This also brings up the other good reason for such studies,in that few researchers can study the long-term effects ofcertain variables, especially when it runs into decades.
-
8/3/2019 Report Jan 24
65/98
y For this study of long-term and subtle effects, theyhave to use pre-existing conditions and medicalrecords. The researcher may want to study anextremely small sample group, so it is easier to startwith known cases and works backwards.
y The thalidomide cases, for example, are an example ofan observational study where researchers had to workbackwards, and establish that the drug was the causeof disabilities.
-
8/3/2019 Report Jan 24
66/98
y The main problem with observational studies is that the
ex
perimenter has no control over the composition of the controlgroups, and cannot randomize the allocation of subjects. Thiscan create bias, and can also mask cause and effectrelationships or, alternatively, suggest correlations where thereare none (error in research).
y For example, in the smoking example, if the researcher foundthat there is a correlation between smoking and increased ratesof lung cancer, without knowing the full and completebackground of the subjects, there is no way of determining
whether other factors were involved, such as diet, occupation orgenetics.
-
8/3/2019 Report Jan 24
67/98
Despite the limitations, an observational studyallows a useful insight into a phenomenon, and
sidesteps the ethical and practical difficulties ofsetting up a large and cumbersome medical researchproject.
-
8/3/2019 Report Jan 24
68/98
Written materialsy
Documents areausefulsourceofdatainqualitativeresearch,buttheyhavetobetreatedwith care. Themostwidelyusedareofficialdocuments,personaldocuments,andquestionnaires.
y Officialdocuments includeregisters,timetables,minutesofmeetings,planningpapers,lessonplansandnotes,
confidentialdocumentsonpupils,schoolhandbooks,newspapersandjournals,schoolrecords,filesandstatistics,noticeboards,exhibitions,officialletters,textbooks,exercisebooks,examinationpapers,work cards,blackboardwork,photographs.
yAnyofthesemightgiveusefulinformation,buttheydonotallprovideanobjectivetruth. Theyhavetobecontextualisedwithinthe circumstancesoftheirconstruction
-
8/3/2019 Report Jan 24
69/98
-
8/3/2019 Report Jan 24
70/98
y
a detailed plan of a scientific ex
periment thatspecifies experimental methods, data collection andsampling schedules.
-
8/3/2019 Report Jan 24
71/98
y statement of purposey materials to be usedy control groups to assess
y effect of the experiment on the tested groupy data interpretation methodsy references to enable readers to understand the
reasoning behind the plan
-
8/3/2019 Report Jan 24
72/98
y This is a formal statement which encompasses your
hypothesis. It is a statement of what question you aretrying to answer and what hypothesis you wish totest.
-
8/3/2019 Report Jan 24
73/98
y List all major items needed to carry out yourexperiment. This list need not be lengthy if thematerials are already published, but it should includethe essentials.
-
8/3/2019 Report Jan 24
74/98
-
8/3/2019 Report Jan 24
75/98
y Identify the relevant control(s) treatment. Thinkabout the variable(s) you and your group are
manipulating. Your control needs to be held undernatural, or unmanipulated conditions, not affected bythe tested variable.
-
8/3/2019 Report Jan 24
76/98
y What will be done with the data once it is collected?Data must be organized and summarized so that thescientist himself, and other researchers candetermine if the hypothesis has been supported or
negated. Results are usually shown in tables andgraphs (figures). Statistic analyses are often made tocompare experimented and controlled populations.
-
8/3/2019 Report Jan 24
77/98
y Any published works (journals, books, websites) thatyou cite in your protocol should be listed in thereference section so that anyone reading your
protocol can look that work up if they desire.
-
8/3/2019 Report Jan 24
78/98
-
8/3/2019 Report Jan 24
79/98
We have made the observation that not all
coffee bean plants mature identically. We
have come up with the followinghypothesis: Nutrient resources in fertilizers
are essential to coffee bean growth, lack
of fertilizer retards growth.
-
8/3/2019 Report Jan 24
80/98
20 coffee bean seeds, soil from a constant
source, fertilizer with a known amount of
nitrogen and phosphorus, pots to plant the
seeds, a constant UV light source.
-
8/3/2019 Report Jan 24
81/98
1.There will be two groups of seeds with 10
plants each: a) the seeds which have
fertilizer (independent variable); and b) the
seeds which do not have fertilizer (control
treatment).
-
8/3/2019 Report Jan 24
82/98
2. Plant all seeds in 30 cm diameter pots
with soil. The fertilizer treatment will receive
10 grams of fertilizer.
3. At the end of the growing season, thenumber of beans,dependent variable, of
each plant will be recorded.
-
8/3/2019 Report Jan 24
83/98
y The plants which have natural (no fertilizer) soil
conditions.
-
8/3/2019 Report Jan 24
84/98
y Ahistogram will be used to plot the results. The averagenumber of beans for each group of plants will be plottedon the Y axis (ordinate) and the treatment group will beplotted on the Xaxis (abscissa). At-test will be performedto determine if the treatment group differs from thecontrol group. If the treatment group produces moreseeds than the control, we can then conclude that thetreatment of fertilizer had an effect and the resource inquestion is limited to plants.
-
8/3/2019 Report Jan 24
85/98
Statistical Treatment Of Datay Statistical treatment of data is essential in order to make use of
the data in the right form. Raw data collection is only one aspectof any experiment; the organization of data is equally important
so that appropriate conclusions can be drawn. This is whatstatistical treatment of data is all about.
y There are many techniques involved in statistics that treat datain the required manner. Statistical treatment of data is essentialin all experiments, whether social, scientific or any other form.
Statistical treatment of data greatly depends on the kind ofexperiment and the desired result from the experiment.
-
8/3/2019 Report Jan 24
86/98
y An important aspect of statistical treatment of data isthe handling of errors. All experiments invariablyproduce errors and noise. Both systematic and random
errors need to be taken into consideration.y Depending on the type of experiment being
performed, Type-I and Type-II errors also need to behandled. These are the cases of false positives and falsenegatives that are important to understand andeliminate in order to make sense from the result of theexperiment.
-
8/3/2019 Report Jan 24
87/98
TREATMENT OF DATA AND
DISTRIBUTIONy Trying to classify data into commonly known patterns is a
tremendous help and is intricately related to statisticaltreatment of data. This is because distributions such as
the normal probability distribution occur very commonly innature that they are the underlying distributions in mostmedical, social and physical experiments.
y Therefore if a given sample size is known to be normallydistributed, then the statistical treatment of data is made easy
for the researcher as he would already have a lot of back uptheory in this aspect. Care should always be taken, however, notto assume all data to be normally distributed, and should alwaysbe confirmed with appropriate testing.
-
8/3/2019 Report Jan 24
88/98
y Statistical treatment of data also involves describing the data.
The best way to do this is through the measures of centraltendencies like mean, median and mode. These help theresearcher explain in short how the data are concentrated.Range, uncertainty and standarddeviation help to understandthe distribution of the data. Therefore two distributions with the
same mean can have wildly different standard deviation, whichshows how well the data points are concentrated around themean.
y Statistical treatment of data is an important aspect of allexperimentation today and a thorough understanding is
necessary to conduct the right experiments with the rightinferences from the data obtained
-
8/3/2019 Report Jan 24
89/98
Mean
y Find the average by adding the values in your datatogether and then dividing by the number of valuesoccurring in the data. For instance, if the values in a
given set of data are 4, 6, 10, 13 and 17, you wouldcalculate the average as follows:
y 4 + 6 + 10 + 13 + 17 = 50
y 50/5 = 10
-
8/3/2019 Report Jan 24
90/98
Median
yFind the median by listing the data inascending or descending order and then
determining the value occurring in themiddle of the data. If the values of a givenset of data are 8, 10 and 13, then the medianis 10.
-
8/3/2019 Report Jan 24
91/98
Mode
yFind the mode by determining thevalue that occurs most often in thedata. If the data values are 2, 4, 5, 6,6, 9 and 10, then the mode is 6.
-
8/3/2019 Report Jan 24
92/98
Standard Deviation1. get the Mean
to begin you need the mean or the average, for exampleadd 23, 92, 46, 55, 63, 94, 77, 38, 84, 26 ... = 598 divideby 10 (the actual number of numbers) 598 divided by10 = 59.8
so the mean or average of 23, 92, 46, 55, 63, 94, 77, 38,84, 26 is
59.8
-
8/3/2019 Report Jan 24
93/98
2. get the deviations
subtract the mean from each of the numbers, theanswers are;-
-36.8, 32.2, -13.8, -4.8, 3.2, 34.2, 17.2, -21.8, 24.2, -33.8
-
8/3/2019 Report Jan 24
94/98
3. square these
to square means multiply them by themselves, theanswers are;-
1354.24, 1036.84, 190.44, 23.04, 10.24, 1169.64, 295.84,
475.24, 585.64, 1142.44
-
8/3/2019 Report Jan 24
95/98
4. add the squares
total of these numbers is 6,283.60
-
8/3/2019 Report Jan 24
96/98
5. divide by total number of numbers less one;-
you had 10 numbers less 1 is 9 numbers
so 6283.60 divided by 9 = 698.18
-
8/3/2019 Report Jan 24
97/98
6. square root of result is Standard Deviation
square root is the number multiplied by itself to get698.18 which is:-
26.4 so 26.4 is the Standard Deviation...
-
8/3/2019 Report Jan 24
98/98
Percentage%= F/N x 100
y Where
%= percentage
F= frequency
N= total number100= constant to get the exact percentage