What Are the 5 Service Quality Gaps? How to Apply SERVQUAL
What Are the 5 Service Quality Gaps?
When you search for “what are the 5 service quality gaps,” the easiest mistake is to treat them as five criteria for scoring a service. The model actually explains why customers may perceive service quality as lower than their initial expectations. Each gap appears at a different point in the chain from customer expectations, management perceptions, service standards, service delivery, and external communication.
The five-gap model is associated with research on service quality by Parasuraman, Zeithaml, and Berry. The SERVQUAL scale is commonly cited as (Parasuraman et al., 1988). In this approach, perceived service quality is formed by the difference between what customers expect and what they perceive after using the service.
The central idea can be expressed as follows:
Perceived service quality = Customers’ actual perceptions - Customers’ expectations
If perceptions are lower than expectations, customers rate the quality as inadequate. If perceptions are equal to or higher than expectations, the quality assessment tends to be more positive. In a quantitative thesis, you usually do not measure all five gaps directly with a single variable. You use the model to explain possible causes, then build a service quality scale and test its relationship with satisfaction, intention to return, or loyalty.
You can also review the SERVQUAL model before deciding on the research model for your topic.
Components of the Model
The five gaps are not five identical observed variables. They describe five points where the process of designing and delivering a service can break down.
| Gap | Position in the service process | Meaning for analysis | Observable indication |
|---|---|---|---|
| Gap 1 | Between customer expectations and management perceptions | Managers do not correctly understand what customers need | Customers value speed, but the business focuses on appearance |
| Gap 2 | Between management perceptions and service standards | The business understands the need but has not translated it into specific standards | There is a service requirement but no clear processing time |
| Gap 3 | Between service standards and the service actually delivered | Employees or systems do not follow the standards correctly | The procedure requires a fast response, but customers wait a long time |
| Gap 4 | Between the service actually delivered and external communication | Advertising or promises create expectations higher than the business can deliver | The website promises “24/7 service,” but support is available only during office hours |
| Gap 5 | Between the service customers expect and the service they perceive | The final result in the customer’s assessment | Customers give a low rating even though the business believes the procedure was followed |
Gap 5 is the part you usually measure with a customer questionnaire. The first four gaps help explain causes within the business or its systems. Therefore, if your study surveys customers using Likert items, state clearly that the data reflect customer perceptions. The data are not sufficient to conclude with certainty that Gap 1, Gap 2, Gap 3, or Gap 4 exists inside the business.
The five gaps are also different from the five dimensions commonly found in the SERVQUAL scale. The five dimensions are reliability, responsiveness, assurance, empathy, and tangibles. These dimensions are aspects used to measure service quality. The five gaps are a framework for explaining where service quality may deteriorate.
The Original Model and Common Extensions
The original model focuses on the difference between expectations and perceptions. Based on this framework, SERVQUAL was developed as a scale containing observed variables that reflect the aspects customers use to evaluate a service. The commonly cited source for SERVQUAL is (Parasuraman et al., 1988).
In applied research, you may see three ways to use the model.
Use the Five-Gap Model as a Theoretical Framework
This approach is suitable when you want to explain how service quality is formed and propose management solutions. You present the five gaps in the theoretical background chapter, then select one or more dimensions to measure with survey data.
For example, a study of banking service quality might argue that the gap between customer expectations and perceptions is an overall expression of problems involving reliability, responsiveness, and empathy.
Use SERVQUAL to Measure Service Quality
You design a questionnaire with an expectations section and a perceptions section, or measure perceptions only, depending on your research design and your supervisor’s guidance. When using SERVQUAL, state the scale version, the service context, and how the wording was adapted for respondents in your study.
The SERVQUAL page can help you check the names of the dimensions, but you still need to review the original scale and your specific research model.
Use SERVPERF to Focus on Perceptions
Another approach measures service quality from actual perceptions instead of asking about expectations separately. This approach is commonly called SERVPERF in service quality studies. You can read more about the SERVPERF model to distinguish between the two questionnaire designs.
Keep the theoretical model, research questions, and data consistent. If you write that the study measures the gap between expectations and perceptions but your questionnaire contains only perception items, the committee may ask how you calculated Gap 5.
Scales Commonly Used for Each Construct
The table below is a reference map for connecting the five-gap framework with research variables. It is not a validated questionnaire for your topic. You need to review the original scale, adapt it to the context, and run the tests on your actual data.
| Construct or component | Content to measure | Example observed variable | Reference source |
|---|---|---|---|
| Reliability | The ability to fulfill promises and provide a stable service | The service provider delivers the service within the promised time | (Parasuraman et al., 1988) |
| Responsiveness | Willingness to support customers and respond to them | Employees respond to requests within an appropriate time | (Parasuraman et al., 1988) |
| Assurance | Knowledge, conduct, and ability to create trust | Employees clearly explain the procedure and emerging issues | (Parasuraman et al., 1988) |
| Empathy | The degree of attention given to customers’ individual needs | Employees pay attention to the customer’s specific circumstances | (Parasuraman et al., 1988) |
| Tangibles | Facilities, equipment, and the visible presentation of the service | The service space, equipment, or interface appears professional | (Parasuraman et al., 1988) |
| Perceived service quality | Overall evaluation after the experience | I rate this provider’s service quality as good | (Parasuraman et al., 1988) |
If you use a Likert scale, state the number of points and the meaning of each point. The Likert scale originated from the attitude measurement technique presented in (Likert, 1932). Each observed variable in the questionnaire should express one idea. Avoid combining “employees are fast and friendly” in one item because respondents may evaluate these two characteristics differently.
After entering the data into SPSS, you can check reliability with Cronbach's Alpha. An alpha of 0.7 or above is generally considered acceptable according to (Nunnally, 1978). For a new scale or exploratory research, an alpha from 0.6 may be considered according to (Hair et al., 2010). Corrected Item-Total Correlation should be 0.3 or above according to (Nunnally and Bernstein, 1994). These are sourced reference criteria, not automatic reasons to delete an item.
Applying the Model to Your Topic
Suppose you are studying food delivery service quality and intention to continue using an app. The model may begin with the five service quality dimensions. Perceived service quality may be a mediator or an overall independent construct, while intention to continue using the app is the dependent variable.
A shortened conceptual model might include:
- Reliability affects perceived service quality.
- Responsiveness affects perceived service quality.
- Assurance affects perceived service quality.
- Empathy and tangibles may also be included if they fit the context.
- Perceived service quality affects intention to continue using the app.
You can write the hypotheses in the following form:
H1: Reliability has a positive effect on the perceived service quality of customers using food delivery applications.
H2: Responsiveness has a positive effect on the perceived service quality of customers using food delivery applications.
H3: Assurance has a positive effect on the perceived service quality of customers using food delivery applications.
H4: Perceived service quality has a positive effect on intention to continue using food delivery applications.
This is an illustrative model, not a conclusion about your study. If the context is a hospital, university, bank, or hotel, the observed variables must reflect the actual experience in that setting. You also need to decide whether to measure expectations and perceptions or perceptions only. These two choices lead to different questionnaire designs and different interpretations of the results.
How to write this in your thesis
You can adapt the wording in your thesis as follows: “Based on the five-gap theoretical framework and the SERVQUAL scale of (Parasuraman et al., 1988), this study proposes examining the effects of [service quality dimensions] on [dependent variable] in the context of [study population].” Replace the bracketed placeholders with the exact variables and context in your model.
How to Test the Model in SPSS or SmartPLS
If your model uses Likert observed variables and you want to test reliability, EFA, correlations, and regression, SPSS is easy to follow. Run Analyze > Scale > Reliability Analysis to check Cronbach's Alpha, then use Analyze > Dimension Reduction > Factor for EFA. In the EFA output, inspect KMO and Bartlett's Test, Total Variance Explained, and Rotated Component Matrix. A KMO of 0.5 or above and a statistically significant Bartlett’s test can be referenced to (Kaiser, 1974). A factor loading of 0.5 or above can be referenced to (Hair et al., 2010).
After EFA, run the regression at Analyze > Regression > Linear. The Coefficients table reports the coefficient, p-value, and VIF. Check the dependent variable, independent variables, coefficient signs, and statistical significance against the significance level you selected in advance. VIF below 5 is a reference threshold in PLS-SEM reporting according to (Hair et al., 2019), but when you use SPSS, you still need to explain exactly how you checked multicollinearity in your study.
SmartPLS is suitable when your model contains latent variables, multiple paths, and a need to assess the measurement model and structural model together. In SmartPLS 4, create a project, import the .csv file, drag the latent variables onto the canvas, connect the paths, and run PLS-SEM Algorithm. Then use Bootstrapping to test the path coefficient and confidence intervals. The procedure should be separated into two parts: assess the measurement model first, then assess the structural model. This follows the two-step approach presented in (Anderson and Gerbing, 1988).
When reading SmartPLS output, you usually inspect Outer Loadings, Construct Reliability and Validity, Discriminant Validity, Path Coefficients, R-Square, f-Square, and Collinearity Statistics. CR of 0.7 or above and AVE of 0.5 or above can be referenced to (Fornell and Larcker, 1981). HTMT below 0.85, or below 0.90 for closely related constructs, is proposed in (Henseler et al., 2015). For bootstrapping, 5,000 subsamples is the practice described in (Hair et al., 2022).
Do not put CFI, TLI, or RMSEA into a SmartPLS report as if they were mandatory PLS-SEM criteria. These indices belong to the fit assessment commonly used in CB-SEM, with the thresholds stated in (Hu and Bentler, 1999). If you use AMOS, the discussion of fit indices is different. If you use SPSS and regression, report the correct regression output instead of combining indices from several methods.
You can review the five service quality gap model to compare the model presentation, but the final results table must come from your .sav or .csv file. An illustrative output table can be presented as follows:
| Hypothesized relationship | Path coefficient | T Statistics | P Values | Illustrative conclusion |
|---|---|---|---|---|
| Reliability → Service quality | 0.286 | 3.412 | 0.001 | H1 supported in the illustrative output |
| Responsiveness → Service quality | 0.174 | 2.106 | 0.035 | H2 supported in the illustrative output |
| Assurance → Service quality | 0.091 | 1.284 | 0.199 | Insufficient evidence to support H3 |
| Service quality → Intention to continue | 0.521 | 7.863 | 0.000 | H4 supported in the illustrative output |
The numbers above are illustrative output, not the results of a real study, and must not be copied into Chapter 4. When you have your output, replace each value with your own result, keep the variable names unchanged, and check the direction of the effect.
Common Mistakes When Using This Model
Calling the Five Gaps Five Independent Variables
Many articles put Gap 1 through Gap 5 into a regression table as five independent variables even though the questionnaire measures only the five SERVQUAL dimensions. Distinguish between an explanatory framework and a variable measured directly. If you do not have data on management perceptions, internal standards, or business communication, you should not make quantitative conclusions about the first four gaps.
Mixing SERVQUAL and SERVPERF in the Same Conclusion
The questionnaire contains both expectations and perceptions, but the method section describes the study as measuring perceptions only. Finalize the design before distributing the questionnaire, then use the correct terminology in Chapters 3 and 4.
Deleting Items Just to Make the Indices Look Better
An increase in Cronbach's Alpha after deleting an item does not by itself prove that the item is theoretically inappropriate. Review the item content, Corrected Item-Total Correlation, factor loading, and the reason for deletion. Record every run so that you can explain your decision when asked.
Renaming Variables After Running the Model
If TC1 and TC2 belong to reliability but you rename one as DU1 in the file or results table, checking the output becomes difficult. Use consistent variable codes from the questionnaire and data file through to the SPSS or SmartPLS output.
Using Another Study as the Only Scale Source
A study in a similar industry can help you understand how to adapt a scale, but you still need to trace the original source and state the context. A scale used for hotels may not retain the same wording when transferred to digital banking or hospitals.
Frequently asked questions
What are the 5 service quality gaps?
They are a framework for explaining five points of difference that may reduce service quality: between customer expectations and management perceptions, between perceptions and standards, between standards and actual delivery, between actual delivery and communication, and finally between customer expectations and perceptions.
Are the 5 gaps in SERVQUAL the 5 scale dimensions?
No. The five gaps are breakdown points in the process of designing and delivering a service. The five dimensions commonly used in SERVQUAL are aspects for measuring service quality: reliability, responsiveness, assurance, empathy, and tangibles.
Can you use SPSS with the five service quality gap model?
Yes, if you design the questionnaire and research variables appropriately. SPSS can support Cronbach's Alpha, EFA, correlation, and regression. If the model has many latent variables and paths, SmartPLS may be more convenient, but choose the tool according to the model and data type rather than choosing it simply because its output contains many tables.
Do you have to measure both expectations and perceptions?
That depends on how you apply SERVQUAL. If you want to calculate the difference between expectations and perceptions, the questionnaire must collect both sections and you must describe the formula clearly. If you measure perceptions only using a SERVPERF approach, report that design accurately and avoid calling the result a gap between expectations and perceptions.
How many observed variables are enough for a SERVQUAL scale?
There is no fixed number for every topic. The number depends on the scale source, service context, wording, and your model. When adapting the scale, preserve the full meaning of the construct, pilot the questionnaire, and test Cronbach's Alpha, EFA, or the measurement model according to your chosen method.
What should you do if Cronbach's Alpha is not acceptable?
Check the coding, reverse-coded items, missing data, and the content of each item before deleting anything. Review Corrected Item-Total Correlation and the Alpha if Item Deleted column, then compare the result with the theory. If you delete an item, record the reason and rerun the related analyses instead of selecting only the option that increases the index.
Open your data file now, compare each observed variable with the five SERVQUAL dimensions, decide whether you are measuring expectations and perceptions or perceptions only, and prepare the variable coding table before running Cronbach's Alpha. If you need to run the analysis on your own .sav or .csv file, you can use DoThesis M4 analysis.