Procedures and Data Collection Methods Back to Class

Procedures and Data Collection Methods Back to Class 11

Procedures (Methodology) ® Probably more attention has been given to this part of research than to any other. It includes experimental design, measurement, statistical analysis and computer programming. It is often the chief section looked at in order to judge whether to fund a proposal. It should contain the following:

Procedures (Methodology) An outline of the overall research design accounting for each objective – variables to be considered, conditions to be controlled and conditions to be eliminated ® Operationalization of the variables, conditions, etc. if this has not already been done. . Usually variables are operationalized in terms of the instruments which are used to measure them. If you are going to use a particular instrument, specify it. If you are going to develop an instrument, state how and give sample items. ®

Procedures (Methodology) You must state how you are going to protect human subjects. Letters you are going to send for approval, statements you will have subjects sign, etc. must be shown. Include how you will obtain informed consent and how anonymity will be preserved. ® Describe data analysis methods. If possible show your data tables will be set up. If you are using statistics, be sure to specify your level of significance (Usually p<. 05), This must be done before the data are collected. ®

Procedures (Methodology) ® Briefly discuss internal and external validity of the design ® State any special conditions in the design that will affect the conclusions or generalizations ® Be sure to state what samples are involved, what are the sources of the data, what processes will be used to gather the data and analyze it, what conditions will be controlled

Procedures (Methodology) ® Pilot studies should be conducted: ® If the technique is unfamiliar to the researcher ® If the instrument is newly constructed ® If the instrument has not been used with this population ® To see if the subjects can handle the instrument ® To give the staff experience in administration and analysis ® As a trial run for the data collection technique ® To provide some data on which to test analysis techniques

Data Collection Methods ® Physiological and ® Five sources ® ® ® Physical Chemical Microbiological Anatomical Observation through the senses Data sources ® ® ® Physical Measures In vivo In vitro Instrument systems ® Subject, stimulator, sensor, signal, display, record

Data Collection Methods ® Use ® ® ® Advantages ® ® Physical outcomes – as criteria against which nursing actions can be assessed Exploration of ways in which nursing actions, including measuring and recording physiological functioning, can be improved objectivity, precision, sensitivity Disadvantages ® ® device may change measurement high energy concentration

Data Collection Methods ® Types Circulatory ® Respiratory ® Neurological ® Muscular-skeletal ® GI function ® GU function ® Glandular function ®

Data Collection Methods ® Observational ® Phenomena Methods amenable to observations Characteristics and conditions of individuals ® Verbal communication behaviors ® Non-verbal communication behaviors ® Activities ® Skill attainment and performance ® Environmental characteristics ®

Data Collection Methods ® Units of analysis – decide what a unit is Molar approach – observe large units of behavior and treat as a whole ® Molecular approach– observe smaller and highly specific behaviors as units ® ® Observer/observed relationship Concealment/no intervention ® No concealment/intervention ® Concealment/intervention ®

Data Collection Methods ® Observational ® methods Unstructured observation l l Participant observation – observer lives in the situation and tries not to interject his views and meanings – takes field notes on l Subjects, setting, behaviors, frequency and duration of events Use of anecdotes in an illustrative fashion Advantages – deeper understanding Disadvantages –observer bias and influence

Data Collection Methods ® Structured observations – look at absence, presence or frequency of a phenomenon ® Categories l l l ® Careful, explicit definitions of behavior or characteristics to be observed No overlapping categories – mutually exclusive Observer interference – no. and skill of observers Checklists l l Talley behaviors (watching called sign analysis) Categorize at regular intervals

Data Collection Methods ® Rating scales l l ® Observer rates some phenomenon in terms of points along a descriptive continuum May be used at intervals or to summarize an entire event Observational sampling l l Time sampling – select a time period during which observation will take place Event sampling – select an event to sample, especially if it is infrequent

Data Collection Methods ® Training observers – even when researcher does most of the research him/herself ® ® ® Advantages ® ® Need a dry run to familiarize with nature of the things to be observed and the tools to be used Results from observers recordings should be compared for inter-rater reliability Captures and directly records behaviors Disadvantages ® Ethical problems, human perceptual errors, demanding of time and emotions, anticipation, hasty decisions

Data Collection Methods ® Interview schedules and questionnaires ® Their format can be anything from rigid standardization to structure absence ® Form of questions Open ended – the subject responds in his/her own words – less bias but hard to analyze ® Close ended – the researcher provides fixed alternatives ®

Data Collection Methods ® Close ended questions cont. l l l Dichotomous – two alternatives Multiple choice – three to five alternatives Cafeteria question – choose responses that most adequately state your view Rank order questions – rank your responses on a continuum from most to least Graphic rating scales – bipolar – specify two opposite ends of a continuum. Respondents give a judgment of something along this ordered dimension

Data Collection Methods ® Question content Facts about the respondent ® Facts about persons known to the respondent ® Facts about events and conditions known ® Beliefs about what the facts are: risk of Ca ® Attitudes, feelings and opinions ® Reasons for or influences on attitudes, etc. ® Level of knowledge about policies, practices, conditions or situations ® Intentions or statements about future actions ®

Data Collection Methods ® How to develop questions Draw up a table of specifications – types of information you are interested in ® Weigh how much emphasis to give each area – this gives you an idea of the number of questions you need to cover the variables ®

Data Collection Methods ® Question wording ® ® Clarity – clarify in your own mind, avoid double-barreled questions, state in lay rather than technical terms if respondents are lay, state in the affirmative Ability of the respondent to reply or give information l l l Use language of the least educated Don’t assume a level of information a person “ought to have” Define technical words Use filter questions – if answer is no, skip to question— Don’t take for granted that a person remembers something even though he/she was present when it happened

Data Collection Methods ® Bias - a serious problem on self-report instruments Assume that the respondent is honest and minimize the bias introduced by the researcher. Don’t suggest answers. Don’t identify a position or attitude with a prestigious group. Avoid emotionally-loaded words. Try to counterbalance the “slant” of questions ® Response sets – things that bias responses such as social desirability, extreme responses ®

Data Collection Methods ® Sensitive or personal information Try to develop more objective wording or offer a close ended question with a range of alternatives ® If dealing with unacceptable behavior – try to create an atmosphere of non-judgment – use alternatives – they are easier to check off than to respond to in an open-ended manner ® Use impersonal wording – not I’m pleased ® Be polite – “Please respond” ask whole sentences – “What is your sex? ” ®

Data Collection Methods ® Response ® Cover alternatives all significant alternatives and usually have an “other – please specify” ® Don’t have overlapping alternatives – 1 -2, 3 -5, 5 or more. ® Place alternatives in some kind of rational order or, if no order, place alphabetically ® Don’t make alternatives too long

Data Collection Methods ® Question sequence Should be psychologically meaningful ® Put open-ended questions first so that respondents give their own opinions before seeing the wording of other questions ® Put demographic information at the end ®

Data Collection Methods ® Format ® Introduction and instructions l l l What is the purpose the researcher is trying to accomplish How did the researcher get the name of the respondent What will be done with the information – confidentiality, anonymity, copies of results Deadline for returning – how to return Researcher’s name and how to reach Answering implies consent

Data Collection Methods ® Format cont. Don’t put too many questions in too small a space ® Set off the alternatives from the stem of the question and align them vertically – ask the respondent to circle, check, or use a separate answer sheet ® Set off the subsets of a filter question so they are not confusing ®

Data Collection Methods ® Steps in construction Make preliminary decisions as to form and the type of information needed. Set up mock tables to see how to analyze ® Draft the questionnaire – monitor the words, look at existing questionnaires, decide on the order of the questions ® Pre-testing and revising – discuss the draft with someone knowledgeable, pre-test to determine clarity ®

Data Collection Methods ® Administration ® of the instrument Questionnaire l l Distribution – the best way is to give it to the whole group at one time, then collect it. Next best is to personally deliver it and/or pick it up. The last resort is to mail it – if the response rate is above 50%, the return is probably sufficient. A return envelope helps. Follow-up reminders – 2 -3 weeks after the first mailing send a letter with a second copy of the questionnaire, or a telephone call is made. Kep a log of incoming receipt of mail on a daily basis

Data Collection Methods ® The interview ® ® ® Put the respondent at ease so he expresses hones opinions – be neat, punctual, courteous, friendly, unbiased and permissive Accept all opinions as natural Don’t read the instrument schedule, but follow the wording precisely If there are many alternatives, hand the respondent a card with them on Don’t paraphrase or summarize the respondent’s reply If he gives only partial answers or “beats around the bush” probe, but do so neutrally – “Explain? ” “Anything else? ”

Data Collection Methods ® Advantages of questionnaires ® ® ® Less costly than interviews More anonymity and no interviewer bias Advantages of interviews ® ® ® Have a high response rate Get people who cannot fill out questionnaires Less ambiguous, deeper Fewer “I don’t knows” Control over the order of presentation Additional non-verbal data can be gathered

Data Collection Methods ® Scales and psychological measures ® Likert scale –several declarative statements expressing a viewpoint on a topic are generated. There should be an equal number of favorable and unfavorable statements - which need to be identified. One concept should have 10 -20 items. ® Originally there were 5 categories of agreementdisagreement, but many people use seven: ® SA – Sl. A - ? – Sl. D – SD. Depending on agreement with the concept being studied, the scoring can be 1 thru 7 or 7 thru 1. No responses are treated as ? or given a value of 4.

Data Collection Methods ® Semantic Differential ® ® A technique to measure the psychological meaning of concepts or objects to an individual. The subject rates a concept on a series of seven point bipolar scales such as fair-unfair, good-bad, importantunimportant, strong-weak, beautiful-ugly, worthlessvaluable, pleasant-unpleasant, cold-warm, responsibleirresponsible, successful-unsuccessful. Adjectives should apply to the concept. Should have some that are evaluative, some related to potency, some related to activity. The pairs should be randomly reversed

Data Collection Methods ® Other scales ® Existing psychological scales – attitude scales, personality measures ® Intelligence, aptitude and achievement tests ® Value scales ® Interest scales

Data Collection Methods ® Content analysis ® An objective, quantitative description of a communication or a document Select variables to be recorded ® Select the unit of content – words, themes, entire items, space/time measure ® Develop a category system for classifying units of content: yes/no or present/absent ® Train coders ®

Data Collection Methods ® Projective techniques ® Pictoral methods – Themic Apperception Test, Rorschach Inkblot Test, ® Verbal methods ® ® ® Expressive methods ® ® Ambiguous verbal stimulus – person is asked to respond with first thing that comes to mind (word association) Sentence completion – elicit attitudes Play technique –drawing, painting, role playing Psychodrama – subjects play themselves Sociodrama – subjects play the part of others Among the most controversial techniques

Data Collection Methods ® Indirect measures ® Letters to the editor, to the Right to Life Association, to the Music Association, etc. ® Responses to vignettes – who is the main character – male/female ® Records and available data

Data Collection Methods ® Q-sort methodology ® Sort a deck of 60 -100 cards according to specified criteria – most like me/least like me or approve/disapprove ® 9 -11 stacks of cards are placed on a table with the number of cards to be in each stack determined by the researcher. The subject is forced to choose where to place the cards. Items put in one place force or influence where others are placed – called a forced choice or ipsative measure (A Likert scale is a normative measure – each item is independent

Data Collection Methods ® least ® 1 approved of most approved of 2 3 4 5 6 7 8 9 category ® 2 4 7 10 14 10 7 number of cards 4 2

Data Collection Methods ® Sociometry ® Information is gained about social choice and interaction patterns of individuals in groups Who do you like, which three people would you like to work with ® Sociograms ®

Data Collection Methods ® Delphi Technique ® Several rounds of questionnaires are sent out. After the first round, the respondents answers, comments, opinions arguments etc. are summarized analyzed and sent to the group along with a redesigned questionnaire. The respondents re-rate, re-vote or re-rank items or make other responses, comments etc. The idea is to obtain consensus and the rounds keep going until it is reached Back to Class 11