STATISTICAL METHOD
Multidimensional Scaling (MDS)
Statistical Consulting for Distance-Based Maps
We build the map for your multidimensional scaling analysis from a dissimilarity that fits your data, choose metric or nonmetric scaling, judge stress against dimensionality and random data, and read relative distances without assigning meaning to the axes.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
METHOD OVERVIEW
What Is Multidimensional Scaling?
Multidimensional scaling represents objects, such as brands, items, species or respondents, as points in a low-dimensional space so that the distances between points approximate their observed dissimilarities. Metric MDS fits the dissimilarity values themselves, while nonmetric MDS fits only their rank order. Stress measures how far the map departs from the data, and it is read with the number of dimensions and objects, not against a fixed cutoff. Axes can be rotated or reflected freely. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides multidimensional scaling analysis within your study design, with the relevant checks, estimates, uncertainty measures, interpretation and publication-ready reporting.
WHEN TO ENGAGE
When Is a Distance-Based Map Appropriate?
MDS fits a question about how objects relate to one another in overall similarity, and these are the research situations in which researchers most often ask us to apply or confirm it.
Perceptual Mapping Respondents rate or compare brands, products or services, and the team wants a map of which are seen as similar, with the limits of the map stated.
Maps From a Dissimilarity Matrix The data are already distances or similarities, such as genetic, ecological or co-occurrence measures, and a visual configuration of the objects is needed.
Metric or Nonmetric Scaling Dissimilarities may be only ordinal, or have ties and missing pairs, so the type of scaling and the number of dimensions must be chosen and justified.
Reviewing a Completed Analysis A two-dimensional map was judged by a stress rule of thumb, and a reviewer has questioned it. Independent Statistical Review reconsiders the dissimilarity and the stress reading.
ANALYTICAL APPROACH
How We Apply Multidimensional Scaling
Within the family of Multivariate, Scale & Reliability, the map is only as sound as the dissimilarity it is built from.
Choosing the Dissimilarity The dissimilarity defines the map, so it must fit the data: Euclidean for scaled variables, Gower for mixed types, or direct similarity ratings. A poor choice creates misleading geometry.
Metric or Nonmetric Fit Nonmetric MDS uses ranks and suits ordinal dissimilarities. Metric MDS suits values that are meaningful as distances. Iterative stress fitting can stop at local minima, so we use many random starts.
Stress and Dimensionality We read stress with a scree-style plot across dimensions and a comparison with maps of random data, not a universal cutoff, and prefer the fewest dimensions that fit and read clearly.
Reading the Map Rotation and reflection are arbitrary, so we interpret relative distances and groupings, not axis directions, and add external variables as fitted vectors only as descriptive aids.
DELIVERABLES
What You Receive
As part of our Quantitative Data Analysis service, the map is delivered with the estimates, checks and files needed to report and defend it.
Data & Variable Review The dissimilarity matrix, its construction, ties, missing pairs, outlying objects and the scaling of the input variables are reviewed, and the measure is documented before any map is fitted.
Method-Specific Checks Stress across dimensions, the comparison with random configurations, the stability of the map across random starts and resamples, and Shepard plots are examined, with sensitivity to the dissimilarity.
Statistical Results Coordinates in the chosen dimensionality, stress values by dimension, a Shepard diagram for the fit, and, where requested, resampling intervals for the positions of the objects.
Publication-Ready Tables & Figures Configuration maps, scree-style stress plots and Shepard diagrams, formatted to APA 7 or the journal's requirements, ready to insert into the manuscript.
Interpretation The map is described through relative distances and groupings of objects, with fit and stability stated, and without naming the axes as constructs or reading the orientation as meaningful.
Reproducible Analysis Files R scripts, Python notebooks, SPSS syntax or Stata do-files are provided where applicable, so the map and its random starts can be rerun if the data change.
ALTERNATIVE METHODS
When Another Method May Be More Appropriate
We choose the method by the input and the question. MDS maps dissimilarities, and these methods fit raw variables, formal groupings, known groups or latent structure.
01
Principal Component Analysis (PCA)
When the data are a case-by-variable table with roughly linear structure, PCA gives interpretable loadings for each variable and component scores for later analysis.
02
Cluster Analysis
When formal groups of similar objects are needed, not a picture of relative distances, clustering assigns membership explicitly and can be paired with a map for display.
03
Exploratory Factor Analysis (EFA)
When the question is about latent dimensions behind rated items, a factor model estimates loadings and shared variance, which a distance map does not.
04
Discriminant Analysis
When each object carries a known group label, a supervised method learns a rule from the labeled objects and reports how accurately it assigns objects it has not seen.
CASE EXAMPLE
How We Work in Practice
An illustration of how a mapping question moves from a planned scaling to a reportable, defensible result.
ILLUSTRATIVE EXAMPLE · BRAND PERCEPTION RESEARCH · UNITED STATES
A market research agency collected similarity ratings among nine meal-kit brands and wanted a perceptual map showing how consumers group them.
The plan, fixed before any map was drawn, used nonmetric MDS on the averaged dissimilarities because the ratings were ordinal, comparing one to four dimensions by stress against maps of random data, with many random starts. Each respondent had rated only a subset of pairs, leaving some pairs thinly rated, so we treated missing pairs as missing and checked the stability of the map by resampling respondents. The agency’s team wanted to name the axes as premium versus convenience, so we reported relative distances and added the brand attribute ratings as fitted vectors, for description only.
Outcome: The agency received the stress by dimension and against random data, the chosen configuration, the respondent-resampling stability check, the attribute vectors, the plots, the scripts and a methods paragraph.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
SOFTWARE
Statistical Software for Distance-Based Maps
Multidimensional scaling is run in the statistical software researchers already use, with syntax or scripts provided where applicable.
RESEARCH DISCIPLINES
Mapping Similarity Across Research Fields
Dissimilarity data arise wherever objects are compared, and each field brings its own measures of similarity, sampling designs and norms for reading a configuration.
A Defined Process for Your Analysis
We review your dataset, research question and design before any work begins, so the scope, deliverables, timeline and fixed project fee are clear from the outset.
01
Share Your Research Materials
Send your dataset, research question, study design and any existing analysis plan or output. We will let you know if anything else is needed to assess the project.
02
Receive Your Assessment
Your statistically trained Research Coordinator reviews the project and confirms the required analysis, deliverables, timeline and fixed project price before work begins.
03
Analysis & Delivery
Once you confirm the scope, an in-house statistician completes the agreed analysis and delivers the outputs, interpretation and reporting files.
RESEARCHER VOICE
What Researchers Say
Feedback from a researcher who wanted the assumptions and interpretation checked against the study design, not only an output.
“I had already started the Multidimensional Scaling (MDS), but I was not confident about the assumptions and interpretation behind the multidimensional scaling (mds) rather than just obtaining an output. The team checked the setup against our study design, explained the decisions in plain language, and returned results we could actually use. The final calculation documented the assumptions clearly enough for the protocol and ethics submission.”
RESEARCH RESPONSIBILITY
Statistical Support Without Taking Over Your Research
We provide the statistical work for your distance-based map, and the study, its meaning and its authorship stay entirely with you.
Agreed Statistical Support We provide the agreed methodological, analytical and statistical reporting support for your distance-based map, within the scope confirmed before work begins.
No Intellectual Contribution OnlineSPSS claims no intellectual contribution to your research. Our work is statistical support, and authorship remains yours.
Your Research Decisions The research question, any meaning given to a dimension, interpretation and conclusions remain with you, and our statistical interpretation stays within what the distances and the design support.
Confidential Handling Research data and project details are handled confidentially and are not published. An NDA is available where required, and can be arranged before sensitive materials are shared.
FREQUENTLY ASKED QUESTIONS
Questions About Stress, Dimensions and Maps
Short answers to what researchers ask before a scaling analysis, from stress values and axes to confidentiality and the assessment process.
What stress value counts as a good MDS fit?
No fixed stress value separates good from bad fits. Stress depends on the number of objects and dimensions, with more dimensions lowering it, and random data produce a nonzero stress as well. We read stress across dimensions, compare it with maps of random data, and weigh how readable and stable the configuration is, so the value is judged in context.
Can I interpret the axes of an MDS map?
Usually not directly. The solution can be rotated or reflected without changing any distance, so the orientation carries no intrinsic meaning. We interpret which objects sit close together or far apart, and where an axis label helps, we support it with external variables fitted to the map and treat it as a description, not a finding.
Should I use MDS or principal component analysis?
Use PCA when the data are a table of continuous variables and interpretable loadings matter. Use MDS when the data are pairwise judgments or distances, such as rated similarity between brands, or when only the rank order of the dissimilarities can be trusted. On Euclidean distances, classical MDS and PCA give the same configuration, so the choice rests on the input and the fit criterion.
Can you review an analysis that has already been completed?
Yes, we can review a completed MDS analysis before submission or a response to reviewers. We check the dissimilarity, the metric or nonmetric choice, the stress and dimensionality decision, the stability of the configuration, the interpretation of the axes and the reporting, and confirm what stands or needs to change.
How do you protect my dataset and research confidentiality?
Your dataset, output files and any protocol or analysis plan are handled confidentially, and project information is not published. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Need Support With a
Distance-Based Map?
Share your research question, distance data and design. We assess whether scaling fits your aim and confirm the approach before any work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014