“SPSS for Introductory Statistics Use and Interpretation 2011” là một cuốn sách giúp sinh viên phân tích và giải thích dữ liệu nghiên cứu bằng cách sử dụng phần mềm IBM SPSS. Cuốn sách này mô tả việc sử dụng thống kê bằng ngôn ngữ dễ hiểu, không chuyên môn để cho người đọc biết cách chọn thống kê phù hợp dựa trên thiết kế, giải thích đầu ra và viết về kết quả. Tác giả giúp người đọc chuẩn bị cho tất cả các bước trong quá trình nghiên cứu, từ thiết kế và thu thập dữ liệu, đến viết về kết quả1. Cuốn sách này là một nguồn tài liệu hữu ích cho những ai muốn tìm hiểu về phần mềm SPSS.
Check the Completed Questionnaires
Review Figures 2.1 and 2.2 carefully to identify any incomplete, unclear, or ambiguous answers Address these issues promptly before moving forward to ensure data consistency Researchers should establish specific rules for handling such responses and document these guidelines on questionnaires or a master coding instructions sheet Consistent application of these rules across all cases is essential for maintaining data quality and accuracy in your analysis.
We identified at least 11 responses across 6 of the 12 questionnaires that require clarification Addressing these responses is essential to ensure data accuracy and validity To resolve these issues, we recommend carefully analyzing each response and documenting specific actions, as illustrated in Figs 2.1 and 2.2 These figures provide visual guidance on how to handle each identified problem effectively, improving the overall quality of the questionnaire data Implementing these corrective measures will enhance response consistency and support reliable research outcomes.
Make Rules About How to Handle These Problems
Establish clear procedures for handling incomplete, blank, unclear, or double answers by creating specific rules prior to data collection whenever possible Consistently applying these rules across all similar responses is essential to prevent bias and ensure the integrity of your results, even when unexpected issues arise during the process.
Interpretation of Problem 2.1 and Fig 2.4
This article explains our approach to handling various data collection issues, emphasizing that alternative choices could have been made The participant data for Participants 1–6 is clear and easily entered using Figure 2.3, while the questionnaires for Participants 7–12 present both minor and significant challenges for data entry We discuss these problems in detail and detail our decisions in numbered callout boxes in Figure 2.4, illustrating the surveys and responses for Subjects 7–12.
In the case of Participant 7, their GPA was recorded as 250, which likely indicates a typo or formatting error, and should be entered as 2.50 While marking it as invalid and coding it as missing is an option, doing so can complicate later data analysis, especially when performing complex statistical procedures Therefore, it is essential to interpret and use the data reasonably whenever possible, ensuring consistency in handling similar data issues throughout the dataset.
For Subject 8, two colleges were examined, and a new legitimate response value (4 = other) was introduced Since the university requires students to be identified with only one of its three colleges, missing values were coded as 98 for cases where multiple colleges were checked or responses did not fit into a single college, such as business engineering or history and business These responses were treated as missing because they appeared invalid or insufficient for analysis Additionally, the code 99 was used for cases with no selections or written responses on the form Using these two distinct codes helped differentiate between types of missing data, ensuring clarity during analysis While some researchers recommend using codes 8 and 9 for missing responses, we prefer employing visually distinctive codes like 98 and 99 for better data management and to avoid confusion with valid codes.
DATA CODING, ENTRY, AND CHECKING 21
Subject 8's GPA is recorded as 2.2, which can be accurately entered as 2.20 for clarity When inputting the GPA, specifying two decimal places ensures the program interprets it correctly as 2.20, maintaining precise and consistent data formatting.
We chose a GPA of 3.00 for Participant 9, based on their statement of "about 3 pt," which suggests their actual GPA could be higher or lower This estimate is considered the most reasonable option given the limited information provided by the student.
Participant 10 only answered the first two questions, resulting in significant missing data and indicating their decision not to complete the questionnaire According to our data quality rule, questionnaires with three or more blank or invalid responses among the first five items are considered invalid and are excluded from analysis; in this case, Participant 10’s questionnaire was not entirely discarded, but it contributed limited data In your research report, clearly state the number of questionnaires excluded and the specific reasons for their exclusion to maintain transparency Typically, no data from invalid questionnaires are entered into the analysis, which means your sample size may be reduced accordingly For example, if a participant leaves the college question blank, we do not delete their entire record but code that particular item as missing or blank.
6 For Subject 11, there are several problems First, she circled both 3 and 4 for the first item; a reasonable decision is to enter the average or midpoint, 3.50
Participant 11 indicated "biology" for their college major, even though this university does not offer a dedicated biology program In such cases, it is appropriate to categorize majors like biology, history, and marketing under the broader arts and sciences category (coded as 1), while majors like civil engineering are better categorized under their specific disciplines (e.g., civil = 3) For unclear or ambiguous examples, refer to the guidelines discussed in Issue 2 to determine the appropriate classification.
Participant 11 entered a GPA of 9.67, which is invalid because the university uses a 4.0 grading system, with 4.00 as the maximum To ensure data accuracy, it’s important to check entered responses for errors; in this case, recognizing that a GPA above 4.0 is invalid allows you to correct the data by leaving the field blank for missing or incorrect information Carefully reviewing completed questionnaires helps identify such errors and maintain the integrity of your dataset.
Enter 1 for Participant 11's reading and homework, despite the boxes being circled rather than checked, as their intent remains clear Additionally, assign a 0 for extra credit, following the standard procedure for all unchecked boxes, except for Subject 10 who did not complete the questionnaire This highlights that the participant’s markings—circles instead of checks—are understood as deliberate indications of their responses.
10 As in Point 6, we decided to enter 2.5 for Participant 12’s X between 2 and 3
11 Participant 12 also left GPA blank so, using the general (system) missing value code, we left it blank
9 Enter 1 for reading and homework
8 For now enter 9.67, but see accompanying discussion
5 Leave all variables blank, except enter 99, missing, for college
Fig 2.4 Completed survey with callout boxes showing how we handled problem responses
DATA CODING, ENTRY, AND CHECKING 23
Clear communication of your rules is essential for effective data handling Ensure that these decisions are visible to all data entrants, such as by highlighting them in callout boxes like those shown in Fig 2.4 A practical approach is to write your guidelines directly on questionnaires, potentially using a different color to enhance visibility and consistency. -Streamline your data rules with clear, visible guidelines—boost accuracy effortlessly like the callout boxes in Fig 2.4! [Learn more](https://pollinations.ai/redirect/draftalpha)
Define and Label the Variables
To begin, create a data file by entering your data into the program If the program isn't open, log in to access it Once the startup window appears, click the "Type in Data" button to open a blank Data Editor, similar to the example shown in Fig 2.5 Ensure that the "Display Commands in the Log" option is checked for better visibility of commands For additional guidance, refer to Appendix A to help you get started with the process.
This section guides you on how to name and properly label variables for your data analysis Following this, we will demonstrate the process of entering data into your dataset To start, we will define and assign labels to the first two variables, which are based on 5-point Likert scale ratings, ensuring clear organization and accurate data collection for effective analysis.
To do this we need to use the Variable View screen Look at the bottom left corner of the Data
To determine whether you are in Data View or Variable View in your software, check which tab is highlighted in white If you are currently in Data View, you can switch to Variable View by following the necessary steps.
To access variable settings in SPSS, click on the Variable View tab located at the bottom left of your screen, which displays a layout similar to Fig 2.5 Alternatively, you can double-click on "var" above the blank column on the far left side of the data sheet to open the Variable View.
Fig 2.5 Blank variable view screen in the data editor
This window displays 11 columns designed for comprehensive variable setup, including fields for variable name, type, width, and number of decimals, enabling precise data entry Users can define variable labels and assign value labels to enhance clarity Additionally, options are available to specify missing values (besides blanks), column width, data alignment (left or right), measurement type, and variable role, ensuring thorough and organized data configuration for accurate analysis.
Define and Label Two Likert-Type Variables
We now begin to enter information to name, label, and define the characteristics of the variables used in this chapter
• Click in the blank box directly under Name in Fig 2.5
• Type recommend in this box Notice the number 1 to the left of this box This indicates that you are entering your first variable 1
Pressing Enter inserts the program's default values for variables; however, it's essential to verify these defaults for accuracy and adjust them as necessary to ensure proper functionality.
Short variable names are now acceptable, though brevity is still recommended for clarity Aside from length considerations, standard naming conventions outlined in Chapter 1, footnote 5, should be followed In this book, instructional steps for SPSS are presented with bullets (e.g., click, highlight), while key terms within SPSS windows are emphasized in bold (e.g., Name) to enhance understanding.
Note that the Type is numeric, Width = 8, Decimals = 2, Label = (blank), Values = None,
Missing = None, Columns = 8, Align = right, Measure = scale, Role = input
In this assignment, we will use the default settings for Type, Width, Columns, and Align On the Variable View screen, the default Type is set to Numeric, indicating that the variable entered will be numeric data While Numeric should be used for numbers, String is suitable for words or letters such as “M” for males and “F” for females; however, it is recommended to avoid using words or letters to facilitate statistical analysis, as recoding them as numbers is often necessary In this book, we will consistently keep the Type as Numeric to streamline data analysis.
We recommend keeping the Width at eight, and keeping the Columns at eight We will always
Align the numbers to the right Sometimes, we will change the settings for the other columns
Now let’s continue with defining and labeling the recommend variable
• For this variable, leave the decimals at 2
To label your items, click on the box under "Label" and type "I recommend course." Using a clear and concise label ensures it displays properly in relevant windows and printouts Labels can be up to 40 characters long, but for optimal readability, it's best to keep them around 20 characters or fewer This helps prevent clutter and makes your outputs easier to read.
In the Values column of Fig 2.5, do the following:
• Click on the word “None” and you will see a small blue box with three dots.
To interpret Likert scale data effectively, click on the three dots to access the settings menu, where you'll see a screen similar to Fig 2.6 Adding value labels for the lower and upper ends of the scale can enhance data clarity, making it easier to understand respondents' responses However, labeling is more critical for nominal or unordered data, while for Likert or other ordered data, labels can be considered optional but still beneficial for accurate interpretation.
• Type 1 in the Value box in Fig 2.6
• Type strongly disagree in the Value Label box Press Add
• Type 5 and strongly agree in the Values and Value Labels boxes Your window should look like Fig 2.6 just before you click on Add for the second time
DATA CODING, ENTRY, AND CHECKING 25
• Leave the cells for the Missing to Measure columns in Fig 2.5 as they currently appear
In research, the variable "Recommend" can function as either an independent (Input) or dependent (Target) variable, so it should be set to "Both" for flexibility Different researchers may code these variables differently based on their study design; for example, if "Recommend" is intended only as an independent variable, it should be coded as "Input." Refer to Figure 2.7 for a visual representation of this coding approach Properly defining the role of "Recommend" ensures clarity and consistency in data analysis and improves the accuracy of research results.
Now let’s define and label the next variable
To add a new variable, click on the blank box under "Name" in Row 2 and enter the variable name without spaces, as spaces are not permitted in variable names However, you can include spaces in labels for better readability This process allows you to efficiently define multiple variables in your dataset while maintaining proper naming conventions for effective data analysis.
• Type workhard in the Name column and press Enter
• Click on the box in Row 2 under Label and type I worked hard in the Label column
• Insert the highest and lowest Values for this variable the same way you did for recommend (1
= strongly disagree and 5 = strongly agree)
Keep all the other columns as they are
Define and Label College and Gender
• Now, select the cell under Name and in Row 3
• Call this third variable college by typing that in the box
To ensure clarity in your data, select the third box under Decimals, as no decimal places are needed since respondents chose only one of three colleges When you click on the Decimals box, up and down arrows will appear, allowing you to easily adjust the number of decimal places by clicking the arrows or double-clicking the box to manually enter your preferred value This method helps maintain precise and clean data presentation for your analysis.
• For the purposes of this variable, select or type 0 as the number of decimals
• Next, click the box under Label to type in the variable label college
• Under Values, click on None and then click on the small blue box with three dots
• In the Value Labels window, type 1 in the Value box, type arts and sciences in the Value Label box
• Then click Add Do the same for 2 = business, 3 = engineering, 98 = other, multiple ans., 99
The Value Labels window should resemble Fig 2.8 just before you click Add for the last time
• Under Measure, click the box that reads Scale
• Click the down arrow and choose Nominal because for this variable the categories are unordered or nominal
Your screen should look like Fig 2.9 just after you click on nominal
• Change Role to Input because college will only be used as an independent variable
To define missing values in your dataset, under the Missing menu, select None and click on the three dots icon Then, choose Discrete Missing Values and enter 98 and 99 in the first two boxes, as shown in Fig 2.10 This step is crucial if you need to specify particular values as missing data codes, ensuring accurate data analysis If you omit this step, the designated values will not be recognized as missing, potentially affecting your results.
In data analysis, a missing cell labeled as None can be misleading, as it doesn't indicate that values like 98 and 99 are missing In this context, None in a column signifies the absence of special missing values, with only blank entries being recognized as missing data Understanding this distinction is crucial for accurate data cleaning and preprocessing.
DATA CODING, ENTRY, AND CHECKING 27
Your Data Editor should now look like Fig 2.11
Fig 2.11 Completed variable view for the first three variables
Now define and label gender similarly to how you did this for college
• First, type the variable Name gender in the next blank row in Fig 2.11
• Click on Decimals to change the decimal places to 0 (zero)
• Now click on Labels and label the variable gender
Display Your Dictionary or Codebook
Once you have defined and labeled your variables, creating a codebook or variable dictionary is essential for documenting your work The codebook serves as a comprehensive record of variable labels and values, complementing the information in the variable view While both contain similar details, the codebook provides a more complete printed reference, ensuring clarity and organization for your data analysis.
• Select File → Display Data File Information → Working File Your codebook should look like Output 2.1, without the callout boxes The codebook is divided into parts: the
Variable Information (which is very similar to the variable view in Fig 2.12) and the Variable Values (which are partially hidden in the variable view)
You may not be able to see all of the file information/codebook on your computer screen However, you should be able to print the entire codebook
DATA CODING, ENTRY, AND CHECKING 29
Missing Values recommend 1 I recommend course
This article emphasizes the importance of academic effort and achievement It highlights that working hard and consistently applying oneself is crucial for success in college Key academic indicators such as GPA, reading, homework, and extra credit completion are vital measures of student performance Properly engaging in these activities can significantly enhance overall academic outcomes and demonstrate dedication to personal growth and success.
Variables in the working file
Value Label recommend 1.00 stongly disagree
5.00 strongly agree workhard 1.00 strongly disagree
5.00 strongly agree college 1 arts & science
4.00 All A's reading 0 not checked/blank
1 check homework 0 not check/blank
These are the value labels for this nominal or unordered variable
These are the labels for the lowest (1), and highest (5) values for the recommend variable
This indicates that 98 and 99 are special/new/missing value codes
These are the values for this dichotomous variable
Most variables use blanks, the system missing value, but college has two missing value codes, 98 and 99
This means the data for this variable will be shown as up to eight digits with two decimal places (See Fig 2.12.)
Enter Data
To access the data editor, close the codebook and click on the Data View tab at the bottom of the screen The spreadsheet displays numbered rows on the left, each corresponding to a study participant Each participant’s questionnaire responses are organized on a single row, with individual columns representing specific variables such as recommend, workhard, college, and others.
After defining and labeling the variables, your next task is to enter the data directly from the questionnaires or from a data entry form
Researchers often transfer questionnaire data to a data entry form before inputting it into SPSS, especially when questionnaires are hard to read, responses come from multiple sources, or additional coding is needed While this method helps prevent direct entry errors, it can also introduce copying mistakes and requires extra time Using a data entry form offers both advantages and disadvantages as an intermediate step To streamline the process, well-prepared questionnaires should allow direct data entry from figures like Fig 2.1 and Fig 2.4 into the data editor, reducing the need for additional steps If challenges arise, researchers may rely on Table 2.1, acknowledging that it adds extra effort to the data entry process.
Table 2.1 presents the data as they would appear if transferred from the cleaned questionnaires to a data entry sheet, whether digitally or handwritten on ruled paper.
Table 2.1 A Data Entry Form: Responses Copied From the Questionnaires
Recommend Workhard College Gender Gpa Reading Homework Extracrd
To enter the data, ensure that your Data Editor is showing
• If it is not already highlighted, click on the far left column, which should say recommend
To input data into the highlighted column, just type the number and press the right arrow key For example, enter '3,' and the number will appear in the blank space above the row, ensuring quick and accurate data entry.
Data coding, entry, and checking involve inputting variable names and corresponding values into the system To do this, select the appropriate variable, press the right arrow key to enter the variable name into the highlighted box, and then input the specific data, such as typing "5" in the workhard column This process ensures accurate data entry and facilitates efficient data management for analysis.
In Fig 2.13, all the data for the participants have been entered
If you click on this button, the value labels instead of the numbers will show
Fig 2.13 Data Editor participants entered
Please input the data from your cleaned questionnaires into Figures 2.1 and 2.4 If you make an error during data entry, simply click on the highlighted cell, correct the score, and press Enter or the arrow key to update the information accurately.
Before you do any analysis, compare the data on your questionnaires with the data in the Data
When managing large datasets, it is preferable to review the entire data set rather than relying solely on a sample Sampling can be useful for initial error detection; however, if errors are found in the sample, a comprehensive check of all entries is essential to ensure data accuracy and integrity.
Run Descriptives and Check the Data
To effectively analyze your data and identify potential errors or issues in your questionnaires, we recommend running the Descriptives statistics program This tool helps you obtain basic descriptive statistics for all your subjects, providing valuable insights into your dataset To do this, follow the necessary steps within the software to ensure accurate and comprehensive results.
• Select Analyze → Descriptive Statistics → Descriptives… (see Fig 2.14) 2
To access descriptive statistics in the software, first open the Analyze menu, then choose Descriptive Statistics from the initial flyout menu, and finally select Descriptives from the subsequent flyout options.
After selecting Descriptives, you will be ready to compute the mean, minimum, and maximum values for all participants or cases on all variables in order to examine the data
• Now highlight all of the variables To highlight, click on the first variable, then hold down the
To select multiple variables in SPSS, hold down the “Shift” key and click on the last variable to highlight all listed variables (see Fig 2.15a) Starting from SPSS version 14, a symbol appears to the left of each variable name, indicating its measurement level—nominal, ordinal, or scale—which is important for proper data analysis Understanding measurement levels is essential for selecting appropriate statistical tests and is discussed in detail in Chapter 3 of this book.
Fig 2.15a Descriptives— before moving variables
• Click on the arrow button pointing right The Descriptives dialog box should now look like Fig 2.15b
DATA CODING, ENTRY, AND CHECKING 33
Fig 2.15b Descriptives— after moving variables
• Be sure that all of the variables have moved out of the left window If your screen looks like Fig 2.15b, then click on Options You will get Fig 2.16
The initial steps involved reviewing key descriptive statistics, including the Mean, Standard Deviation, Minimum, and Maximum values The Standard Deviation was then deselected, indicating that no additional descriptive statistics are needed at this stage These analyses will be further expanded upon in Chapter 4, where more detailed statistical procedures will be addressed.
Ensure the Variable list bubble is selected in the Display Order section for proper arrangement You can choose between Ascending or Descending order to organize your variables accordingly For an alphabetically sorted list, simply check the Alphabetic option.
• Click on Continue, which will bring you back to the main Descriptives dialog box (Fig 2.15b)
• Then click on OK to run the program
You should get an output like Fig 2.17 If it looks similar, you have done the steps correctly
The syntax or log is a valuable tool for verifying your requests and executing or rerunning advanced statistical analyses If the syntax does not appear in your output, refer to Appendix A for guidance.
Fig 2.17 Output viewer for Descriptives
The left side of Fig 2.17 displays the different components of your output, allowing you to customize each part for better clarity and presentation By clicking on items such as Title, Notes, or Descriptive Statistics, you can activate and edit those sections directly For instance, selecting the Title enables you to expand it or add relevant information like your name and date, enhancing the professionalism of your report For detailed instructions on editing outputs, refer to Appendix A.
To customize your output, double-click on the large, bold "Descriptives" label in Fig 2.17 and enter your name in the prompted box, ensuring your name appears on the printed results Additionally, type "Output 2.2" at the top of the output window to clearly identify the specific output for future reference or review by your instructor This process enhances the clarity and organization of your statistical analysis results.
When analyzing each variable, it is essential to compare the minimum and maximum scores from Fig 2.17 with the highest and lowest acceptable values outlined in the codebook (Output 2.1) This data verification step helps ensure accuracy by detecting potential data entry errors and confirming the correct application of missing data codes Conducting this initial check safeguards the integrity of subsequent statistical analyses and promotes reliable research outcomes.
Certainly! Please provide the article you'd like me to rewrite with SEO compliance.
DATA CODING, ENTRY, AND CHECKING 35
DESCRIPTIVES VARIABLES=recommend workhard college gender gpa reading homework extracrd
/STATISTICS=MEAN STDDEV MIN MAX
I worked hard college gender grade point average
Average college is not meaningful
The number of people with no missing data
This analysis presents the number of participants (N) with complete data for each of the eight variables, with a total of nine participants (Valid N) having no missing information across all variables The table includes the minimum and maximum scores observed for each variable, illustrating data range and variability For instance, no participants selected a score of 1 on any variable, while some chose a 2 for the "I recommend" course, and at least one participant scored a 5 Notably, the variable "I worked hard" has both its minimum and maximum scores at 5, indicating it is a constant rather than a variable and unlikely to contribute meaningful data to subsequent statistical analyses.
The table also provides the Mean or average score for each variable Notice the mean for I worked hard is 5 because everyone circled 5 The mean of 1.80 for college, a nominal
The mean value of 0.55 for dichotomous variables such as gender, reading habits, and homework completion indicates that approximately 55% of respondents identified as female and answered "yes" to reading and homework questions The reported average GPA of 3.58 appears unusually high for most undergraduate groups and likely contains an error Notably, a maximum GPA of 9.67 is impossible at this university, which has a maximum GPA of 4.00; hence, the score of 9.67 for participant 11 is invalid It is essential to review the questionnaires for potential data entry errors, and any erroneous values like 9.67 should be replaced with the missing value code or left blank to ensure data accuracy.
2.1 What steps or actions should be taken after you collect data and before you run the analyses aimed at answering your research questions or testing your research hypotheses?
When considering data coding rules for questionnaires, it is essential to ensure clarity and consistency to enhance data quality and analysis accuracy Additional rules should address standardized coding for ambiguous or open-ended responses to reduce variability Existing rules may require modification to incorporate recent best practices, such as implementing digital coding methods or increasing transparency in coding procedures By updating and expanding these guidelines, researchers can improve data reliability, facilitate reproducibility, and streamline data management processes Clear, comprehensive coding rules are vital for producing valid, comparable results across studies.
2.3 Why would you print a codebook or dictionary?
2.4 If you identified other problems with the completed questionnaires, what were they? How did you decide to handle the problems and why?
If the university permits double majors across different colleges, it is essential to implement clear policies to manage students enrolled in multiple colleges simultaneously When both colleges are checked, the system should verify that students meet the specific requirements of each program and coordinate credit transfer and scheduling accordingly Handling such cases involves establishing a unified advising process to ensure students fulfill all graduation criteria, promoting academic flexibility while maintaining academic standards This approach benefits students by allowing interdisciplinary studies and broadening their academic experience, aligning with the university’s commitment to fostering diverse educational opportunities.
It is essential to review raw questionnaire data both before and after entering it into the data editor to ensure accuracy and data integrity Checking data prior to entry helps identify and correct errors, inconsistencies, or incomplete responses, reducing the risk of propagating inaccuracies in the analysis Methods to verify data before entry include thorough manual reviews, cross-checking with original responses, and implementing validation rules during data collection After entering the data, it is important to perform data validation and consistency checks within the data editor, such as using built-in validation tools, spot checks, and statistical summaries, to detect and correct any discrepancies or entry errors, thereby ensuring reliable results in your research.