Sunteți pe pagina 1din 10

Chapter 15 Data Preparation and Analysis Strategy

True/False Questions
1. According to the text, the data preparation process begins with checking the questionnaires for completeness. True (moderate, page !5" The final data analysis plan should never differ from the preliminary plan for data analysis. False (moderate, page !5" !ata preparation should only begin once all batches of questionnaires are received from the field. "n this way, processing uniformity is boosted. False (di##i$ult, page !5" !ata cleaning is the last stage of the data preparation process. False (moderate, page !%" A review of the questionnaires with the ob%ective of increasing accuracy and precision is called editing. True (moderate, page !%" 'oding consists of screening questionnaires to identify illegible, incomplete, inconsistent, or ambiguous responses. False (moderate, page !&" !eleting the question from the survey is an alternative for the treatment of unsatisfactory responses. False (moderate, page !&" *nsatisfactory responses from collected data may be discarded when the proportion of unsatisfactory respondents is 2+ percent to + percent. False (moderate, page !&" The assignment of a symbol to represent a specific response to a specific question, along with the data record and column position that symbol will occupy, is called coding. True (moderate, page !&" According to the text, the symbols used in the assignment of codes to each possible response to each question is usually a number. True (moderate, page !&" A book containing coding instructions and the necessary information about variables in the data set is called a codebook. True (moderate, page !&" -nly software programs developed for the research industry should be used to enter data, not programs such as .icrosoft /0'/1. False (moderate, page !'" A code in which the number of records for each respondent is the same and the same data appear in the same columns for all respondents is called a fixed field code. True (moderate, page !(" 'oding unstructured questions is relatively simple, since the response options are predetermined. False (moderate, page !("
1$)

2. .

#. $. &.

(. ).

,.

1+.

11.

12.

1 .

1#.

1$. 1&. 1(. 1).

According to the text, category codes should be collectively exhaustive and mutually exclusive. True (di##i$ult, page 1!" !ata should be coded in a lean way in order to avoid retaining much detail. False (di##i$ult, page 1!" 'ategories are collectively exhaustive if each response fits into one and only one category code. False (moderate, page 1!" According to the text, categories are mutually exclusive if every response fits into one of the assigned category codes. False (moderate, page 1!" Transcribing data involves transferring the coded data from the questionnaires or coding sheets onto disks or directly into computers by key punching or other means. True (moderate, page 11" -ptical scanning involves direct machine reading of the codes and simultaneous transcription. True (moderate, page 11" The thorough and extensive check for consistency and treatment of missing responses is called validating. False (moderate, page 1)" A part of the data cleaning process that identifies data that are out of range or logically inconsistent, or that have extreme values, is called consistency checks. True (moderate, page 1)" .issing responses represent values of a variable that are unknown, either because respondents provided ambiguous answers or their answers were not properly recorded. True (moderate, page 1*" Treatment of missing responses poses problems, particularly if the proportion of missing responses is between $ percent and 1+ percent. False (moderate, page 1*" 'asewise deletion is an option available for the treatment of missing responses. True (moderate, page 1*" A method for handling missing responses in which respondents with any missing responses are discarded from the analysis is called pairwise deletion. False (moderate, page 1*" 'asewise deletion is a method of handling missing values in which for each calculation or analysis, only the respondents with complete responses are considered. False (moderate, page 1*" 2airwise deletion may be appropriate when the sample si3e is large. True (moderate, page 1*" 4tatistical techniques can be broadly classified as univariate or multivariate. True (moderate, page 15" 4tatistical techniques appropriate for analy3ing data when there is a single measurement of each element in the sample are called univariate techniques. True (moderate, page 15"
1$,

1,.

2+. 21. 22.

2 .

2#.

2$. 2&.

2(.

2). 2,. +.

1.

.ultivariate statistical techniques are appropriate for analy3ing data if there are several measurements of each element and each variable is analy3ed in isolation. False (moderate, page 15" *nivariate techniques are statistical techniques suitable for analy3ing data when there are two or more measurements on each element and the variables are analy3ed simultaneously. False (moderate, page 15" .ultivariate statistical techniques are concerned with the simultaneous relationships among two or more phenomena. True (moderate, page 15" 'onsidering the known characteristics of the data is a stage in selecting a data analysis strategy. True (easy, page 15" A good rule of thumb about the discarding of respondents is that such decisions should be made before conducting any analysis. False (moderate, page 1&"

2.

#. $.

+ultiple Choi$e
&. "n the marketing research process, data preparation and analysis comes immediately after 55555555 and before 555555555555. a6 problem definition7 research design b6 research design7 field work c6 field work7 research design d6 problem definition7 report preparation e" #ield ,or-. report preparation and presentation (easy, page ! " According to the text, the data preparation process begins with 55555555555555. a6 validating the questionnaires b6 evaluating the interviewers c6 pretesting the questionnaire d6 transcribing data e" $he$-ing the /uestionnaires #or $ompleteness (di##i$ult, page !5" 8hich of the following is the last stage of the data preparation process9 a6 questionnaire checking b6 coding c6 data cleaning d6 transcribing e" sele$ting a data analysis strategy (moderate, page !%" A review of the questionnaires with the ob%ective of increasing accuracy and precision is called 55555555555555. a0 editing (moderate, page !%" b. coding c. validating d. pretesting e. transcribing

(.

).

,.

1&+

#+.

55555555555555 consists of screening questionnaires to identify illegible, incomplete, inconsistent, or ambiguous responses. a. 'oding 10 2diting (moderate, page !%" c. :alidating d. 2retesting e. Transcribing 8hich of the following is ;-T an alternative for the treatment of unsatisfactory responses9 a. return the questionnaire to the field to get better data b. assign missing values c. discard unsatisfactory respondents d0 delete the /uestion #rom the sur3ey (di##i$ult, page !%" e. c and d 8hich of the following is ;-T a stage of the data preparation process9 a0 /uestionnaire design (moderate, page !%" b. editing c. transcribing d. data cleaning e. coding Assigning missing values to unsatisfactory responses may be desirable in all of the following situations /0'/2T< a. if the number of responses is small. b. if the proportion of unsatisfactory responses for each of these respondents is small. c. if the variables with unsatisfactory responses are not key variables. d0 All o# the mentioned situations /uali#y #or the assignment o# missing 3alues ,hen unsatis#a$tory responses are dis$o3ered on a /uestionnaire0 (di##i$ult, page !&" e. c and d !iscarding unsatisfactory responses from collected data may have merit in all of the following situations /0'/2T< a. when the proportion of unsatisfactory respondents is small. b. when the sample si3e is large. $0 ,hen the unsatis#a$tory respondents di##er #rom satis#a$tory respondents in o13ious ,ays0 (di##i$ult, page !&" d. when the proportion of unsatisfactory responses for each of these respondents is large. e. responses on key variables are missing. *nsatisfactory responses from collected data may be desirable when the proportion of unsatisfactory respondents is 55555555555555. a0 less than 1! per$ent (moderate, page !&" b. 1+ percent to 2+ percent c. 21 percent to + percent d. 1 percent to #+ percent e. more than #1 percent The assignment of a symbol to represent a specific response to a specific question, along with the data record and column position that symbol will occupy, is called 55555555555555. a. editing 10 $oding (moderate, page !&" c. sampling d. pretesting e. selecting
1&1

#1.

#2.

# .

##.

#$.

#&.

#(.

According to the text, the symbol used in the assignment of codes to each possible response to each question is usually a 55555555555555. a. letter of the alphabet b. =reek symbol $0 num1er (moderate, page !&" d. binary digit e. ampersand A book containing coding instructions and the necessary information about variables in the data set is called a>n6 55555555555555. a. editing book b. sampling record $0 $ode1oo- (moderate, page !&" d. questionnaire e. checking book 8hich of the following is ;-T mentioned in the text as an additional code that could be included for each respondent in the coding process9 a. the pro%ect code b. the interviewer code $0 the super3isor $ode (di##i$ult, pages !54 !%" d. the validation code e. codes for responses to questions A code in which the number of records for each respondent is the same, and the same data appear in the same columns for all respondents is called a 555555555555555. a. variable field code 10 #i5ed #ield $ode (moderate, page !(" c. random code d. virtual code e. uniform code 'oding 55555555555555 questions is relatively simple since the response options are predetermined. a. unstructured b. open?ended $0 stru$tured (easy, page !(" d. free?flowing e. uniform According to the text, category codes should be 555555555555555 and 55555555555555. a. !ependent7collectively exhaustive 10 $olle$ti3ely e5hausti3e. mutually e5$lusi3e (di##i$ult, page 1!" c. mutually inclusive7 dependent d. dependent7 mutually exclusive e. precoded7 post?coded 'ategories are 55555555555555 if each response fits into one and only one category code. a. dependent b. mutually inclusive c. collectively exhaustive d0 mutually e5$lusi3e (moderate, page 1!" e. uniform

#).

#,.

$+.

$1.

$2.

$ .

1&2

$#.

According to the text, categories are 55555555555555 if every response fits into one of the assigned category codes. a. dependent b. mutually inclusive $0 $olle$ti3ely e5hausti3e (moderate, page 1!" d. mutually exclusive e. uniform 55555555555555 data involves transferring the coded data from the questionnaires or coding sheets onto disks or directly into computers by key punching or other means. a. :alidating b. /diting c. 'oding d0 Trans$ri1ing (moderate, page 11" e. 4electing 5555555555555 involves direct machine reading of the codes and simultaneous transcription. a0 6pti$al s$anning (moderate, page 11" b. @ey punching c. :alidating d. !ata cleaning e. 2redictive dialing !ata are verified as they are collected when all of the following methods are used /0'/2T< a. 'AT". b. 'A2T. c. electronic methods. d0 -eypun$hing0 (moderate, page 11" e. c and d The thorough and extensive check for consistency and treatment of missing responses is called 55555555555555. a. validating b. transcribing $0 data $leaning (moderate, page 1)" d. key punching e. editing A part of the data cleaning process that identifies data that are out of range or logically inconsistent, or that have extreme values, is called 55555555555555. a. transcribing b. validating $0 $onsisten$y $he$-s (moderate, page 1)" d. coding e. editing All of the following computer packages can be programmed to identify out?of?range values for each variable and print out the respondent code, variable code, variable name, record number, column number, and out?of?range value /0'/2T< a. .";"TAA. b. /0'/1. c. 4244. d0 768D0 (moderate, page 1)" e. b and d
1&

$$.

$&.

$(.

$).

$,.

&+.

&1.

8hich of the following is ;-T a technique for executing data transcribing9 a. mark?sense forms 10 -eypun$hing c. optical scanning d. 'AT"B'A2" e. ob%ective sort and entry >moderate, page #126 55555555555555 represent values of a variable that are unknown, either because respondents provided ambiguous answers or their answers were not properly recorded. a0 +issing responses (moderate, page 1*" b. "nconsistent responses c. -ut?of?range values d. "nvalid responses e. /dits Treatment of missing responses poses problems, particularly if the proportion of missing responses is 55555555555555. a. between 1 percent and percent b. between percent and $ percent c. between $ percent and ( percent d. between ) percent and , percent e0 more than 1! per$ent (moderate, page 1*" 8hich of the following is ;-T an option available for the treatment of missing responses9 a. substitute a neutral value b. casewise deletion $0 modi#y the /uestionnaire (di##i$ult, page 1*" d. pairwise deletion e. a and c A method for handling missing responses in which respondents with any missing responses are discarded from the analysis is called 55555555555555. a0 $ase,ise deletion (moderate, page 1*" b. pairwise deletion c. alternate deletion d. concurrent deletion e. both a and b 55555555555555 is a method of handling missing values in which for each calculation or analysis, only the respondents with complete responses are considered. a. 'asewise deletion 10 Pair,ise deletion (moderate, page 1*" c. Alternate deletion d. 'oncurrent deletion e. b and d 2airwise deletion may be appropriate when all of the following conditions are present /0'/2T< a. the sample si3e is large. b. there are few missing responses. c. the variables are not highly related. d0 the sample si9e is small0 (di##i$ult, page 1*" e. c and d

&2.

& .

&#.

&$.

&&.

&(.

1&#

&).

55555555555555 are statistical techniques suitable for analy3ing data when there are two or more measurements on each element and the variables are analy3ed simultaneously. a. Transformational displacement b. Transform and reform c. Transform and modify d. :ariable streamlining e0 :aria1le respe$i#i$ation (moderate, page 1 " "f there are too few respondents in the lowest income category, this category may be combined or merged with the next lowest category. This is known as 555555555555. a. recombining b. replication c. returning d. resetting e0 re$oding (di##i$ult, page 1 " 4tatistical techniques can be broadly classified as 55555555555555 or 55555555555555. a. univariate or random b. random or parallel c. parallel or multivariate d0 uni3ariate or multi3ariate (moderate, page 15" e. uniform and non?uniform 4tatistical techniques appropriate for analy3ing data when there is a single measurement of each element in the sample are called 55555555555555. a0 uni3ariate te$hni/ues (di##i$ult, page 15" b. multivariate techniques c. random techniques d. parallel techniques e. uniform techniques 8hich of the following statistical techniques is appropriate for analy3ing data if there are several measurements of each element and each variable is analy3ed in isolation9 a0 uni3ariate te$hni/ues (di##i$ult, page 15" b. multivariate techniques c. random techniques d. parallel techniques e. non?uniform techniques 55555555555555 are statistical techniques suitable for analy3ing data when there are two or more measurements on each element and the variables are analy3ed simultaneously. a. *nivariate techniques 10 +ulti3ariate te$hni/ues (di##i$ult, page 15" c. Candom techniques d. 2arallel techniques e. *niform techniques 8hich of the following statistical techniques are concerned with the simultaneous relationships among two or more phenomena9 a. univariate techniques 10 multi3ariate te$hni/ues (di##i$ult, page 15" c. random techniques d. parallel techniques e. uniform techniques
1&$

&,.

(+.

(1.

(2.

( .

(#.

($.

According to the text, a survey discloses that 1 percent of the respondents preferred the clientDs brand, while +.) per cent preferred the closest competing brand. "n such a case, it is perfectly fine to announce that the clientDs brand is the dominant brand. 4uch a situation would be 555555555555555. a. acceptable based on the facts of the survey b. acceptable is only the client was told this c. impossible to weigh in with a moral %udgment based on these details d. unacceptable if only the client was told this e0 una$$epta1le 1e$ause it is misleading0 The 1rands ha3e 3irtual parity0 (moderate, page 1&"

2ssay Questions
(&. "n a short essay, discuss the seven steps in the data preparation process. 8hen should the data preparation process begin9 8hy is it important to begin the data preparation process at this step9 Ans,er The entire data preparation process is guided by a preliminary plan of data analysis that was formed in the research design phase. As the first step, the field work supervisor checks for acceptable questionnaires. /diting, coding, and transcribing the data, all done by the data collection agency as part of field work, follow this. The researcher cleans the data then performs further checks for consistency and specifies how missing responses will be treated. The researcher then selects an appropriate data analysis strategy. The final data analysis strategy differs from the preliminary plan of data analysis due to the information and insights gained since the preliminary plan was formulated. !ata preparation should begin as soon as the first batch of questionnaires is received from the field, while the field work is still going on. Thus, if any problems are detected, the field work can be modified to incorporate corrective action. (di##i$ult, pages !54 !%" ((. "n a short essay, describe the process of editing in the data preparation process and discuss the four types of errors a researcher is attempting to minimi3e in this procedure. Ans,er 2diting involves reviewing questionnaires to increase accuracy and precision. "t consists of screening questionnaires to identify illegi1le, in$omplete, in$onsistent, or am1iguous responses. Cesponses may be illegible if they have been poorly recorded, such as answers to unstructured or open?ended questions. 1ikewise, questionnaires may be incomplete to varying degrees. A few or many questions may be unanswered. At this stage, the researcher makes a preliminary check for consistency. A response is ambiguous if, for example, the respondent has circled both # and $ on a (?point scale. (easy, page !%" (). "n a short essay, describe the process of coding and discuss the purpose of a field and a record. "nclude a specific example to support your answer. ;ext, describe the purpose of a codebook in the coding process. Ans,er Coding means assigning a code, usually a number, to each possible response to each question. The code includes an indication of the column position and data record it will occupy. Eor example, gender of respondents may be coded as 1 for females and 2 for males. A #ield represents a single item of data, such as gender of the respondent. A re$ord consists of related fields, such as sex, marital status, age, household si3e, and occupation. Thus, each record can have several columns. =enerally, all the data for a respondent will be stored on a single record, although a number of records may be used for each respondent. "t is often helpful to prepare a $ode1oo- containing the coding instructions and the necessary information about the variables in the data set. (moderate, page !&"

1&&

(,.

"n a short essay, define data cleaning and discuss how consistency checks and treatment of missing responses relate to the data cleaning process. Ans,er !ata cleaning is the thorough and extensive checking for consistency and treatment of missing responses. This cleaning process includes consistency checks and treatment of missing responses. 8hile preliminary consistency checks have been made during editing, the checks at this stage are more thorough and extensive, since these are made by computer. Consisten$y $he$-s are a part of the data cleaning process that identify data that are out of range or logically inconsistent, or that have extreme values. !ata with values not defined by the coding scheme are inadmissible. +issing responses represent values of a variable that are unknown, either because respondents provided ambiguous answers or their answers were not properly recorded. 2roper selection, training, and supervision of field workers should minimi3e the incidence of missing responses. (moderate, pages 1)4 1 "

)+.

"n a short essay, list and discuss the three options available to the researcher for the treatment of missing responses on a questionnaire. Ans,er a. 4ubstitute a neutral value F a neutral value, typically the mean response to the variable, is substituted for missing responses. Thus, the mean of the variable remains unchanged and other statistics, such as correlation, may not be affected much if the proportion of missing values is small. b. 'asewise deletion F in casewise deletion, cases or respondents with any missing responses are discarded from the analysis. Gowever, casewise deletion could lead to a significant reduction in the sample si3e and should be used with caution. c. 2airwise deletion F in pairwise deletion, instead of discarding all cases with any missing values, the researcher uses only the cases or respondents with complete responses for each calculation. Gowever, pairwise deletion could also lead to a reduction in the sample si3e. (moderate, pages 1*4 1 "

)1.

"n a short essay, discuss the ethical issues related to data preparation and analysis. Ans,er /thical issues can arise during the data preparation and analysis step, particularly in areas where the researcher has to exercise %udgment. 8hile checking, editing coding, transcribing, and cleaning, the researcher can get some idea about the quality of the data. 4ometimes it is easy to identify respondents who did not take the questionnaire seriously or who otherwise provided data of questionable quality. 'onsider, for example, a respondent who checks the Hneither agree nor disagreeH response to all the + items measuring attitude toward spectator sports. !ecisions about whether such respondents should be discarded, that is, not included in the analysis, can raise ethical concerns. A good rule of thumb is to make such decisions during the data preparation phase before conducting any analysis. !iscarding respondents after analy3ing the data raises ethical concerns, particularly if this information is not fully disclosed in the written report. .oreover, the procedure used to identify unsatisfactory respondents and the number of respondents discarded should be clearly disclosed. (moderate, pages 1&"

1&(

S-ar putea să vă placă și