COMPTRS IN COMM RES
COMPTRS IN COMM RES COM 5331
Popular in Course
Popular in Communication
This 17 page Class Notes was uploaded by Kenny Ziemann on Thursday September 17, 2015. The Class Notes belongs to COM 5331 at Florida State University taught by Barry Sapolsky in Fall. Since its upload, it has received 48 views. For similar materials see /class/205627/com-5331-florida-state-university in Communication at Florida State University.
Reviews for COMPTRS IN COMM RES
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 09/17/15
CHAPTER 26 RESEARCH WRITING Understanding the data analysis process is critical to survey research Often clients will request nothing more than tables derived from frequencies and crosstabulations from the researcher In some instances statistical analyses will be included However it is not uncommon for a client to request and pay for a detailed report in which the researcher describes the survey findings and methodology This can be a timeconsuming taskthat requires attention to detail and above all else strong writing ability Post secondary education in the US has increasingly overlooked the need for writing assignments and critical evaluation of writing Many graduate students lack command ofthe language and an understanding ofthe rules of grammar Students develop bad writing habits and most troublesome seem unaware of the flaws in their writing Yet employers look for strong writing skills in prospective employees Certainly anyone who wishes to pursue a career in survey and market research should develop a solid ability to write It would be a daunting task to include a thorough treatment of how to write in this reader and course Therefore what follows are some examples of bad writing and guidelines for writing research reports Much of this information has been drawn from other sources However no amount of tips can make you a good writer There is no substitute for practice and for critical review of your writing by experienced writers who can give you helpful feedback Part 1 Examples of Writing and Interpretation Problems The following examples are drawn from past student writing in COM5331 Grammatical problems that are all too common in such writing are discussed later In the first part below direct quotes from student writing are used to show the sometimes humorous effects of word choice The second part outlines common problems that appear in student writing Use this as a guide to what to avoid in your own writing The following are taken verbatim from past students39 writing Sometimes you need to listen to what you are saying In a similar vein more than one quarter of VCR owners who rent tapes for entertainment agree that when it comes to owning a VCR renting tapes is important As opposed to quotin a different arteryquot Young subscribers those 18 to 34 years old and older subscribers those 35 and up are split almost in half with 51 and 49 subscribing to premium channels respectively Were they split with an ax chainsaw or what As seen in Table 1 the Democratic students were split down the middle with one third in favor You can imagine the blood in Table 1 The respondents were broken into categories of race white Ouchl There seems to be a tendency for a more widespread belief in death afterlife among lower class Americans Believe it or not I39d bet on it The students are distributed between nearly half 43 who are high news consumers and Like mail Respondents who attend religious services up to several times a week had the highest percentage of attendance These high religious respondents had the What were they smoking We can predict that respondents who find there are poor people in this country because of loose morals and drunkenness are related to attendance of religious services First cousin aunt uncle or what According to the GSS survey tabulations concerning respondents39 religious service attendance and their beliefs that loose morals and drunkenness contribute to reasons for being poor the following information stands out Where In the field think loose morals and drunkenness as a very important reason for the existence of poverty At the same time only a third Orwas it much later while less than one in five 18 indicated that they were not important for explaining poverty Meanwhile ofthose respondents who never attend church back at the ranch At the extremes 5 of respondents believe that their TV viewing has increased over time while 5 believe that viewing has decreased over time Of what Temperature The results of this procedure did not produce any revolutionary data Che Guevera would have been disappointed White AfricanAmerican and Hispanic all had their majorities in favor of gun control On the other side of the coin they are equally similar with a little less than It must be a rather large coin A phrase one may encounter is quotOn the one hand quot which should be coupled with quotOn the other handquot If you must use this phrasing use both parts To get to the heart of this matter the researcher wondered if the self identification of the students as liberalconservative would shed light on More appropriate for a novel or instruction manual on dissection The number of nonwhite respondents that stated they were unsure about affirmative action stayed nearly the same yet those that favored or opposed it changed drastically Watch out for overstatement As with the breakdown in race less respondents were unsure oftheir own opinions on this matter Usually cars break down Be cautioned because again our sample size of Black women is small Usually a yellow blinking light will do In general the table reveals a stainvay pattern members of each social class attend religious services more frequently than members of quotlowerquot classes Going up or down Therefore it is safe to saythat the greater opposition to a woman s right to have an abortion When is it not safe to say this The studies of Jones 1986 and Smith 1987 contain several methodological flaws which were improved upon in Quayle39s 1988 study created even better flaws A slightly different type of study was undertaken that studied identical of course no two studies will be What these examples above should tell you is that your words can have unintended even humorous meaning You need to reread your sentences and think about what they are saying The following list provides useful suggestions for improving research writing For numbers one through nine spell out the word For numbers 10 and higher use numerals uReport percentages as whole numbers Double check and make sure that you have the percentage right If a table shows 52 then the text should also say 52 not 53 or 51 Avoid simple errors Proofreading is not optional it is essential For clarity refer to the survey question when discussing the analysis of a variable At a minimum refer to the question by number eg Q 9 or quote enough of the question so the reader will understand what is being discussed uWhen you are referring to the literal words in a response set put them in quotes For example quotOne quarter 24 ofthe respondents quotstrongly agreequot that quot Don39t interpret percentages based upon a very small count number of cases You simply cannot make statements about cells in a table with few respondents You should request quotcountsquot when running a Crosstabulation or a Tables procedure You may later exclude the counts when generating a quotreport qualityquot table Don39t inject your personal opinions or values into your interpretation of the findings Also do not write in the first person singular quotI believe Avoid the second person voice quotAs you can seequot This reader has varied the voice from quotyouquot to quotwequot to quotonequot If you are citing percentage or other figures from a table in the report make that clear Use wording such as quotTable1 shows Avoid the obvious For example when discussing a dichotomous variable if an overwhelmingly majority gave one response then obviously a very small number gave the alternative response Use the word quotrelationshipquot very carefully The degree of relationship between two variables refers to quotcorrelationquotSometimes the word quotassociationquot is also used to refer to correlation A significant Chi Square value is interpreted as meaning two variables are not independent they are related Therefore unless you have conducted the appropriate statistical test do not use words such as relationship or association Use words that carry very specific meanings correctly or don39t use them at all If you don39t know what a quotp valuequot is what a quotdependent variablequot is or what a quotsignificant differencequot is then don39t use these terms better yet lookthem up in a book on research design and statistics and learn to use them correctly quotMediaquot and quotdataquot are plural nouns quotMediumquot and quotdatumquot are singular This is very basic and something every student in Communication should know uGet your facts straight For instance to say that the 1991 GSS survey respondents are students is simply wrong It is a national survey of adults A researcher must be precise and accurate uCite percentages in the text The statement quotNearly half of the respondents agreed that should read quotNearly half ofthe respondents 48 agreed that However simply citing the percentages in parentheses can lead to incomplete sentences For example this sentence is incorrect quotIn 1999 13 of respondents quotstrongly agreedquot that sexual relations between samesex adults is wrongquot You should be able to remove a parenthetical from a sentence and the sentence should still make sense In this example if you remove the parenthetical 13 the sentence does not make sense In a crosstab table focus on the appropriate percentages and counts Do not discuss the expected values or residuals Also do not discuss the marginals Use Frequencies if you want to report the distribution of responses to a variable the entire sample or subsample if you use Select Cases art 2 Percentages fractions and meanings Some suggestions for wording When you are trying to describe a certain percentage of respondentsresponses here are some suggestions for words you might use to refer to those percentages Range 0110 11 20 What you could write Simple Fraction Onetenth Onefifth Onefourth a quarter three tenths Onethird twofifths Nearly half Half one out of two More than half threefifths Twothirds seven tenths Threefourths fourfifths Ninetenths More than nine out often What you could write Size as a Group Negligible very few a handful Small few Minority Large minority plurality iftrue Very large minority large plurality Evenly divided if close to 5050 Majority most people Large majority substantial considerable Very large majority great Overwhelming very great Nearly everyone almost all Developed by Backstrom amp HurshCesar Survey Research 1981 Part 3 Tips for Writing Reports The following tips were developed by Backstrom and HurshCesar Survey Research 1981 Items in bold are particularly relevant to COM5331 1 Use Fractions Percentages can be intimidating particularly if strung out in a long series If close to a simple fraction use it Instead of saying 65 percent of the women and 32 percent ofthe men sa About twothirds of the women and onethird ofthe men Of course stringing together several fractions can be confusing too 2 Put percentages into parentheses For the above example About twothirds ofthe women 65 and onethird of the men 32 said 3 Use simple ratios Where possible use ratios for comparisons Women by a 2to1 margin over men 65 to 32 were more likely to or Women were twice as likely as men to 4 Think in whole numbers Same example Two out of three women 65 as compared with only one out of every three men 32 said 5 Use simple fractions It s safe to onehalf onethird onefourth and onefifth one out of five to express data Beyond these understanding becomes difficult People don t think in terms of oneseventh 14 As noted in item 7 quottenthsquot can be used effectively to compare fractions 6 Use sensible fractions It s a bit silly to report on a sixteenth ofthe Catholics 6 and twentyfourth ofthe Jews 4 Keep it simple and sensible Only a handful of Catholics 6 and Jews 4 said Use comparable fractions It s confusing to mix fractions We go as far as mixing onefourth ofthe boys and onehalf of the girls But otherwise we try to keep consistency in fractions onetenth of married people as compared with threetenths of the singles and seven tenths of those divorced Use common referents It s also confusing to mix referents as in fiveeighths ofthe Republicans seventenths ofthe Democrats and threefourth ofthe Independents In this instance where the numbers are close it might be more helpful to use a common point as just about threefourths of the Independents 76 and slightly smaller groups of Republicans 72 and Democrats 70 said Don t trivialize statistics In the text we shouldn t report every single statistic we get In particular stringing together a series of percentages can impede understanding The answers were definitely yes 17 probably yes 25 probably no 27 definitely no 14 no opinionnot sure 8 intermediatemiscellaneous response 5 refused to answer 3 not asked the question 00 H Don t trivialize responses Item 9 also is an example of overreporting of categories We get the important information we need from knowing definitely yes 17 probably yes 25 probably no 27 and definitely no 14 We might simply report that people divided about evenly between yes 42 and no 41 Be sensible about small values Of course it is important to report negligible numbers as it is to cite big numbers Only 4 percent felt But further analysis may be silly This 4 percent divided into people who were favorable 3 or who had no opinion 1 Report whole percentages Surveys suffer from errors due to sampling Saying Democrats 453 and Republicans 226 is a false precision Instead we report whole figures rounding off the decimal places to the nearest whole number As a rule of thumb round numbers up 445 45 Verbalize table highlights Tables should not stand alone Many readers cannot or will not read tables The text should highlight important findings and provide the supporting table nearby Use the term significant exclusively We don t use the word significant to mean important since it has a narrow technical meaning in statistical studies Describe the sample In the introduction we indicate the total sample size and the population from which it is drawn Refer readers to the detailed sample description in the appendix Admit to small numbers We warn readers when numbers get small Blue collar workers under age 20 although there were only eight in the sample were all opposed to the gasoline tax Write simply For the lay reader a lengthy blurb on say multiple regression is statistical gobbledygook This virtually assures that busy decision makers will ignore our findings We could say We tried to predict people s reactions to the event based on our knowledge of a combination oftheir characteristics Be descriptive Describe relationships don tjust symbolize them About half ofthe differences in the scores is accounted for by the respondents age older people being better informed than younger people Not We found r 72 which means that age explains onehalf of the variance r2 49 in scores Be consistent in tense A rule ofthumb is to write using the tense formula respondents weredata are This means that we used past tense when describing what people said in response to our questions findings but use present tense when stating the meaning of the findings conclusions Example Fourfifths of the respondents 81 approved the referendum It appears that the measure will pass handily Avoid overly artistic language A vocabulary of 2500 words is enough to write simple declarative sentences for most clients research needs Straining for artistic colorful language can be an obstacle for readers understanding The worst case could be phraseology that suggests vive la difference when the real finding is consistency and similarity Avoid absolutes It s misleading to say The public feels when in fact we mean that only a large proportion of our respondents expressed a certain feeling Surveys look for similarities and differences among people We don t expect to find too many absolutes in human behavior Distinguish intensity and extensity A common mistake is to report that there is strong support among Model City residents 85 for this issue when we mean there is widespread support How strongly intensity each individual supports the issue is different from the number extensity of people who support it Be approximate No figure is exact in a survey Rather than pretending great precision we report data as being close to It s one reason we use fractions to convey a proper humility about social data roughly half ofthe people or about two out of three Don t be surprised One person s surprise is another s naivete The kinds ofjudgments we avoid are those that suggest we have some stake in the outcome such as It was surprising to find or Remarkably few people said We are not supposed to be surprised only to find evidence that helps us accept or reject our hypotheses Report the population lfthe sample is representative of the population we should report the population Model City those voters and avoid such sampleoriented words as respondents interviewees surveyed and the like We use such terms only when we are skeptical of whether those studied do represent the population they re supposed to 26 Reprint the question Although a questionnaire is included in the appendix we often reprint the entire question with the appropriate table in the text in order that readers know exactly what was asked This enables them to check our interpretation 27 Write to communicate Many writers write for themselves rather than for their clients Much of this writing is self conscious overy artistic complex lofty We follow an old advertising dictum in our reports If the Smiths can get it the Smythes will get it too Part 4 A Sampling of Common Writing Problems Grammar and Usage In general students are encouraged to check out William Strunk s book The Elements of Style at wwwbartlebycom1 41 indexhtm uSemicolon and Colon lftwo or more clauses grammatically complete and not joined by a conjunction and are to form a single compound sentence the proper mark of punctuation is a semicolon In other words if each portion of the sentence on either side ofthe semicolon could stand alone as a simple sentence then the semi colon is appropriate The colon signals the reader that the words following it will explain or illustrate a point that was just made It introduces lists examples etc It is also used to separate introductory words from a long or formal quotation which follows A comma would be used to introduce a short quotation Commas Commas are overused and misused There are many rules and only a review of a grammar book can provide a complete overview Hyphenation The tendency today is to avoid using hyphens whenever possible There are however some situations in which hyphens should be used The first one below is the one most ignored by student writers Two or more words modifying a noun etc and used as a single adjective should have a hyphen We encounter this often in certain areas of Communication For example computerbased technology Hyphens are also appropriate in for the following Compound nouns compactdiscinteractive Compound words containing self ex half or quarter as the first element quarterhour Capital letterjoined to a noun Qsort Elements of an improvised compound manonthestreet interviews facetoface communication Compound numerals twentyone through ninetynine Fractions fourfifths uApostrophe An apostrophe can indicate the possessive the communicators are talking to each other but not reacting to each other s message The apostrophe is also used for contractions you re You should also learn the rules for nouns that end in s Complete Sentences Use complete sentences Avoid sentence fragments incomplete sentences Read a sentence out loud in isolation from surrounding sentence Does it communicate a complete thought at a minimum a noun and verb are present uPronounnounverb agreement One area that students consistently have problems with is agreement between nouns pronouns and verbs Some of the following is taken from httpwwwe r 39 n mum u m D I 1 I RN 5 Flamrum Lu Nounverb agreement 1 The number ofthe noun singular or plural determines the form of the verb since verbs must agree with their nouns plural Some readers consider Kohlberg39s theory of moral development problematic because he only examines ways of thinking and not how emotions and gender socialization affect moral development singular Carol Gilligan in her book entitled In a Different Voice challenges the premise of Kohlberg39s theory 2 There may be more than one nounverb pair in a sentence you need to make sure that each nounverb pair agrees in number Elizabeth KublerRoss in her book On Death and Dying outlines the stages of dying and she illustrates that both the terminally ill person and the person39s loved ones experience these stages 3 Here is one less experienced writers often miss The number of shocks increases with the level of annoyance The noun is singular Therefore the verb must reflect this Nounpronoun agreement Nouns and pronouns need to agree on two things Number singularone or pluralmore than one Gender male female neutral 1 For example the following sentences do not make sense since the pronouns do not agree with their nouns in number 1st sentence or gender 2nd sentence Elvis sightings have occurred more abundantly in the last two years he has been occurring at the rate of ten per month I know a woman who likes Elvis Presley39s music so much he trained her dog named Elvis to thump her tail and bark in rhythm to all of its tapes 2 The sentences do make sense when the pronoun gender and number is straightened out Elvis sightings have occurred more abundantly in the last two years they have been occurring at the rate of ten per month quotSightingsquot is the noun to which the pronoun refers it is plural and thus requires the plural pronoun quottheyquot to make sense Note that the verb changes as well since verbs have to agree with their nouns or pronouns I know a woman who likes Elvis Presley39s music so much she trained her dog named Elvis to thump his tail and bark in rhythm to all of her tapes quotWomanquot is a feminine noun so it requires the feminine pronoun quotherquot You can assume because of its name that the dog is male so it would be more correct to say quothisquot tail If you don39t know the dog39s gender you could say quotitsquot tail Yet quotitsquot does not make sense when referring to the tapes since the neutral pronoun quotitsquot implies that the tapes belong to the dog So you could say quother tapesquot to show that they belong to the woman or quothis tapesquot to refer to the tapes of Elvis39 singing Tense Shifting verb tense can confuse the reader Normally do not shift tense However there is often confusion about discussing past research findings A particular research study obtained past tense certain results But a theory used in that study continues to explain thus the theory predicts present tense In general past tense indicates actions or conditions that occurred at a specific point in the past Confusing pairs of words Clarify your understanding of the difference between quotaffectquot and quoteffectquot As a verb affect means quotto influencequot or quotto assumequot As a noun it means quotfeelingquot or quotemotionquot Effect as a verb means quotto causequot and as a noun means quotresultquot Other pairs you should clarify include quotmayquot vs quotcanquot and quotthanquot vs quotthenquot uMiscellaneous suggestions Provide the source reference for direct quotes specific facts figures or specific research findings Don39t overuse quotes lndent long quotes Don39t have a paragraph simply be a quote Also don39t have a paragraph be simply a string of quotes Use consistent form location underlining indentation for headings subheadings lists etc On lists don39t use quotbulletsquot o in one place letters a in another place and numbers 1 in yet another place Avoid beginning sentences with pronouns quotthisquot quottheyquot andquotitquot For example quotHumantohuman communication models may provide the theory underlying interactivity However it is illdefinedquot In the second sentence what does quotitquot refer to Avoid sounding repetitive This can mean using the same word twice in a sentence or in successive sentences It can also mean introducing two or three paragraphs in a row with the same phrasing quotAccording to Avoid overblown claims The worst quotThis is the first study to look atquot And don39t overuse underlining italics or bold art 5 Punctuation Exercise Put in semicolons colons dashes quotation marks Italics use an underline and parentheses where ever they are needed in the following sentences U FFIONT I The men in question Harold Keene Jim Peterson and Gerald Greene deserve awards Several countries participated in the airlift ltaly Belgium France and Luxembourg Only one course was open to us surrender said the eXmajor and we did Judge Carswell later to be nominated for the Supreme Court had ruled against civil rights In last week39s New Yorker one of my favorite magazines I enjoyed reading Leland39s article How Not to Go Camping Yes Jim said I39ll be home by ten There was only one thing to do study till dawn Montaigne wrote the following A wise man never loses anything if he has himself The following are the primary colors red blue and yellow 0 Arriving on the 8 to plane were Liz Brooks my old roommate her husband and Tim their son 1 When the teacher commented that her spelling was poor Lynn replied All the members of my family are poor spellers Why not me He used the phrase you know so often that I finally said No I don39t know The automobile dealer handled three makes of cars Volkswagens Porsches and Mercedes Benz Though Phil said he would arrive on the 9 19 flight he came instead on the 10 36 flight Whoever thought said Helen that Jack would be elected class president ln baseball a show boat is a man who shows off The minister quoted Isaiah 5 21 in last Sunday39s sermon There was a very interesting article entitled The New Rage for Folk Singing in last Sunday39s New York Times newspaper 9 Whoever is elected secretary of the club Ashley or Chandra or Aisha must be prepared to do a great deal of work said Jumita the previous secretary 0 Darwin39s On the Origin of Species 1859 caused a great controversy when it appeared Answers U FFIONT SOQONO The men in question Harold Keene Jim Peterson and Gerald Greene deserve awards Several countries participated in the airlift ltaly Belgium France and Luxembourg quotOnly one course was open to us surrenderquot said the eXmajor quotand we didquot Judge Carswelllater to be nominated for the Supreme Courthad ruled against civil rights In last week39s New Yorker one of my favorite magazines I enjoyed reading Leland39s article quotHow Not to Go Campingquot quotYesquotJim said quotI39ll be home by tenquot There was only one thing to dostudy till dawn Montaigne wrote the following quotA wise man never loses anything if he has himselfquot The following are the primary colors red blue and yellow 10 Arriving on the 810 plane were Liz Brooks my old roommate her husband and Tim their son 11 When the teacher commented that her spelling was poor Lynn replied quotAll the members of my family are poor spellers Why not mequot 12 He used the phrase quotyou knowquot so often that I finally said quotNo I don39t knowquot 13 The automobile dealer handled three makes of cars Volkswagens Porsches and Mercedes Benz 14 Though Phil said he would arrive on the 919 flight he came instead on the 1036 flight 15 quotWhoever thoughtquot said Helen quotthat Jack would be elected class presidentquot 16 In baseball a quotshow boatquot is a man who shows off 17 The minister quoted lsaiah 521 in last Sunday39s sermon 18 There was a very interesting article entitled quotThe New Rage for Folk Singingquot in last Sunday39s New York Times newspaper 19 quotWhoever is elected secretary ofthe clubAshley or Chandra or Aishamust be prepared to do a great deal of workquot said Jumita the previous secretary I O Darwin39s On the Origin of Species 1859 caused a great controversy when it appeared Copyright 19952001 by OWL at Purdue University and Purdue University Before trying this exercise you can review the Purdue online Writing Lab39s handout Semicolons Colons Dashes Quotation Marks and Italics at httn39lowl ennli h niirrlue overvwhtm Part 6 Review a Report for Readability After writing a survey report review it for readability The conventional wisdom is that most people are comfortable reading below their actual level For general audiences ease of reading is especially important Here is one formula 1 Take a 100word sample ofthe survey report 2 Compute the average number of words in each sentence lfthe final sentence in your sample runs beyond 100 words use the total number of words at the end ofthat sentence to compute the average 3 Count the number of words in the 100word sample with more than two syllables Do not count proper nouns or threeword syllable forms ending in ed or es 4 Add the average number of words per sentence to the number of words containing more than two syllables and multiply the sum by 04 Example Suppose a 100word passage contains an average of 20 words per sentence and 10 words of more than two syllables The sum ofthese is 30 Multiplying 30 by 04 gives you a score of 12 This means that the passage requires a 12thgrade reading level Part 7 Judging the Comprehensiveness and Accuracy of a Survey Research Report The following checklist can be used to assess the quality of a research report Many ofthe questions relate to concepts that go beyond the scope of this course The checklist should convey to you the high standards that good research and good research reporting must meet Introduction and Background 1 Are the survey39s main objectives or guiding questions stated measurably 001 lfthe survey is part of a research study are the research questions or hypotheses stated precisely Is a description given of how the survey fits into the context of previous surveys done locally andor elsewhere Are the people or agencies that a commissioned the survey and b conducted the survey acknowledged Are the actual writers of the report acknowledged Is an explanation provided of the problem or need the survey39s data are to resolve Survey Content Is the total number of questions given Is the content of the survey adequately described Is the number of questions given for each scale or subscale Is the content described adequately for each scale Are the response choices adequately described Is the time for administration specified Is the time needed for individuals to complete the survey given Is the relevant literature included Design and Sampling Is the design described adequately Is the design justified If a sample are sampling methods adequately described If a sample are the survey39s participants randomly selected If more than one group are the survey39s participants randomly assigned lfthe unit that is sample eg students or employees is not the population of main concern eg teachers or managers is this addressed in the report eg in the analysis or discussion If a sample and a nonrandom sampling method is used is evidence given regarding the similarity of the groups at baseline lf groups are not equivalent at baseline is this problem adequately addressed in the analysis or the interpretation Are criteria given for including all sampling units eg students teachers and whoever else is studied Are criteria given for excluding units Is the sample size justified say with a power calculation ls information given on the number of participants in the source population Is information given on the number of participants a eligible to participate b who agreed to participate c who refused to participate d who dropped out or were lost to followup before completing the survey ls information given on the number of respondents a who completed all questions b on whom some data are missing If observations or measures are made over time is the time period justified Are reasons given for a individuals or groups who dropped out b for missing data Are the effects on generalizability of choice equivalence and participation of the resultant sample explained Are the effects on internal validity of choice equivalence and participation of the resultant sample explained Reliability and Validity Are the independent variables defined Are the dependent variables defined Are data provided on the survey39s reliability for each variable Are data provided on the survey39s validity for each variable Are the methods for ensuring reliability eg quality assurance and training described Are the methods for ensuring reliability adequate Are the methods for ensuring validity described Are the methods for ensuring validity adequate Are the scoring methods adequately described 42 Are the scaling methods described 43 Are the scaling methods adequate 44 Is the survey39s administration adequately described 45 Is information provided on methods for ensuring the quality of data collection 46 Is the duration of the survey justified 47 Is the duration sufficient for the survey39s objectives 48 Are the effects on the survey39s generalizability and practicality of the selection reliability validity of data sources and the length of data collection explained Data Analysis 49 Are statistical methods adequately described 50 Are statistical methodsjustified 51 Is the purpose of the analysis clear 52 Are scoring systems described 53 Are potential cofounders adequately controlled for in the analysis 54 Are analytic specifications of the independent and dependent variables consistent with the survey39s research questions or hyptheses 55 Is the unit of analysis specified clearly Reporting 56 Are references given for complex statistical methods 57 Are complex statistical methods described in the appendix 58 Are exact 12 values given 59 Are the results of the analysis clearly described 61 Are the survey39s findings clearly described 62 Do the conclusions follow from the survey39s results 63 Are the survey39s limitations discussed adequately 64 Does the validity of the findings outweigh the limitations Part 8 Just for fun Understanding and writing scientific research papers The following list of phrases and their definitions might help you understand the mysterious language of science and medicine These special phrases are also applicable to anyone working on a PhD dissertation or academic paper anywhere answers to the published any sense required before a complete understanding ofthis are the experiment and to Cindy Adams for valuable me what it meant Part 9 Writing Example In a national survey respondents 2003 Multicultural Study were asked to mark where they place themselves on the attitudinal behavioral spectrum of optimismpessimism 7point scale with pessimism 1 and optimism 7 The following CROSSTABS table looks at the relationship between optimismpessimism and opinion on dependence on technology A Chisquare analysis was also requested How might you describe the key findings for this analysis In your description it is important to observe the writing guidelines in this chapter Below the tables is an example of such a description Use this example as a guide as you do Exercise 11 TABLE 1 Level of OptimismPessimism and View on Dependence on Technology OPTHVIISTICPESSHVIISTIC Pessimistic Middle Optimistic 13 Ground 45 67 Total OUR SOCIETY 1 STRONGLY DISAGREE 13 31 46 90 HAS BECOME 30 26 33 30 TOO DEPENDENT ON TECHNOLOGY 2 DISAGREE 85 269 330 684 194 222 238 225 3 AGREE 214 633 701 1548 487 521 505 509 4 STRONGLY AGREE 127 281 312 720 289 231 225 237 Total 439 1214 1389 3042 1000 1000 1000 1000 ChiSquare Tests Asymp Sig Value df 2sidedL Pearson ChiSquare 11167a 6 083 Likelihood Ratio 10931 6 091 Linearb Linear Association 739420 1 39006 N of Valid Cases 3042 a 0 cells 0 have expected count less than 5 The minimum expected count is 1299 Table 1 examines the relationship between the optimism scale and views on society s dependence on technology The optimism scale was reduced to three groups pessimists optimists and those who consider themselves neither pessimists or optimists In general pessimists and optimists alike agree that society is too dependent on technology approximately onehalf of each group Pessimists were slightly more likely to strongly agree 29 that society has become too dependent on technology compared to 23 of optimists According to the Chisquare analysis the two variables are not related X2 1117df 6 p 083 Here is a breakdown ofthe above paragraph 1 Table 1 examines the relationship between the optimism scale and views on society s dependence on technology The first sentence provides context forthe analysis Here the context is very limited In a research report the writer must explain the source of the data and the specific questionsscales used The question number could be referred to as in Q 7 2 The optimism scale was reduced to three groups pessimists optimists and those who consider themselves neither pessimists or optimists This sentence makes it clear that the original optimismpessimism scale was recoded or collapsed to create three responses low medium and high Notice that the word RECODE is not used as the writer stays away from using the jargon of SPSS 3 In general pessimists and optimists alike agree that society is too dependent on technology approximately onehalf of each group A look at the table shows that the percentages when read horizontally are in general fairly close That is whether one is a pessimist a middlegrounder or an optimist agreement or disagreement with the technology statement is fairly similar If residuals were included in the table they would be uniformly small What is evident is that about onehalf of each group on the optimism scale is in agreement with the technology statement percent who agree 4 Pessimists were slightly more likely to strongly agree 29 that society has become too dependent on technology compared to 23 of optimists The largest spread between percentages appears in the row for strongly agree If residuals were present the cells for this row would have the largest residuals Moreover this row shows a large spread between the extremes on the optimism scale the low scorers vs the high scorers These percentages stand out among all the cells ofthe table Do not discuss the marginals the row or column totals 5 According to the Chisquare analysis the two variables are not related X2 1117 df 6 p 083 A Chisquare analysis was performed The last sentence explains that the test result shows the two variables are independent The proper way to cite the Chisquare results is shown in parentheses This should be used exactly as it appears whenever citing the Chisquare analysis results Exercise 11 a Use the data file GSSQ3SAV Write a paragraph describing the results ofthe crosstabulation analysis below Discuss what you believe to be the key percentages Finally comment on the nature ofthe relationship between these two variables LLOK BIRTH CONTROL T0 TEENAGERS 1416 PARTNER2 Number of Sexual Partners in Past Year Crosstabulatiol PARTNER2 Number of Sexual Partners in Past Year 200 2 or 00 None 100 One more Total PILLOK BIRTH 1 STRONGLY AGREE Count 43 156 44 243 CONTROL To Expected Count 504 1638 287 2430 TEENAGERS 1415 within PARTNER2 Number ofSexual 231 258 415 271 Partners in Past Year Residual 74 78 153 2 AGREE Count 45 195 35 275 Expected Count 571 1854 325 2750 within PARTNER2 Number of Sexual 242 323 330 307 Partners in Past Year Residual 121 96 25 3 DISAGREE Count 53 146 18 217 Expected Count 450 1463 257 2170 within PARTNER2 Number of Sexual 285 242 170 242 Partners in Past Year Residual 80 3 77 4 STRONGLY DISAGREE Count 45 107 9 161 Expected Count 334 1085 190 1610 within PARTNER2 Number of Sexual 242 177 85 180 Partners in Past Year Residual 116 15 100 Total Count 186 604 106 896 Expected Count 1860 6040 1060 8960 within PARTNER2 m er of Sexual 1000 1000 1000 1000 Partners in Past Year ChiSquare Tests Asymp Sig Value df 2sided Pearson ChiSquare 25838a 6 000 Likelihood Ratio 26048 6 000 Linearb Linear Y 20 777 1 000 AssOCIatIon N of Valid Cases 896 3 0 cells 0 have expected count less than 5 The minimum expected count is 1905 NOTE Syntax use for this analysis RECODE partners 00 11 2 thru 72 INTO PARTNER2 EXECUTE b Using the data file STUDOZSAV Write a paragraph describing the results of this analysis Discuss what you believe to be the key percentages Finally comment on the nature ofthe relationship between these two variables V14A Women in this country have equal job opportunities with men V34 What is your sex Crosstabulation V34 What is your sex 1 Male 2 Female Total V14A Women in this 100 Agree Count 83 70 153 country ve equal j b Expected Count 685 845 1530 opportunities With men Within V34 What is your sex 430 294 355 Residual 145 145 200 Neutral Count 36 31 67 Expected Count 300 370 670 Sitgtt isnyzifsex 187 130 155 Residual 60 60 300 Disagree Count 74 137 211 Expected Count 945 1165 2110 Sitgtt isnyzifsex 383 576 490 Residual 205 205 Total Count 193 238 431 Expected Count 1930 2380 4310 WithinV34 o o o Whatis oursex 1000 1000 1000 ChiSquare Tests Asymp Sig Value df 2sided Pearson ChiSquare 15762a 2 000 Likelihood Ratio 15873 2 000 LinearIbyLinear 13854 1 000 AssOCIatIon N of Valid Cases 431 3 0 cells 0 have expected count less than 5 The minimum expected count is 3000 Syntax used for the above Crosstabulation RECODE v1 32 1thru21 4thru 53 INTO v14a EXECUTE
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'