ES EN
Monte Carlo Research Lab · Soft Exclusion Coding Protocol
Final empirical results and documented internal review of soft-exclusion mechanisms across the Monaco fieldwork corpus.
This page presents the final empirical application of the Soft Exclusion Coding Protocol (SECP) to the consolidated Monaco Research Observatory corpus. The results are based on final coding decisions reviewed by the researcher responsible, concerning observation opportunities specific to each case and category, rather than using all 75 cases as the universal denominator for every mechanism.
Review Status
Internally reviewed · Final occurrence coding
The internal review of SECP occurrence decisions has been completed for this version of the corpus. The results are final and distinguish between observable opportunity, confirmed occurrence, reviewed absence and situations that are uncertain or not assessable.
On this page, validation refers to operational checking of decisions against the protocol rules and documented evidence. The review was conducted by a single coder: no intercoder agreement was estimated, and no comprehensive independent validation of the instrument is claimed.
Data release: 75 spatial cases · 225 visual assets · 12 SECP categories · 900 case×category assessments.
Page Structure
Contents
- Empirical Corpus
- Fieldwork Design & Dataset
- Unit of Assessment
- Observation Opportunities
- Normalization Rule
- Occurrence-Level Coding
- Internal Review & Reliability Scope
- Final Empirical Results
- Comparative Reading
- Negative Cases & Interpretive Controls
- Empirical Interpretation
- VAM–SECP Relationship
- Interpretative Limits
- Validation Scope & Conclusion
Empirical Corpus
Consolidated observational evidence base
Visual assets provide evidence for assessment, but asset ≠ occurrence. A SECP occurrence may be supported by one or more assets without being multiplied by the number of available photographs or videos.
Fieldwork Design & Dataset
Empirical design, traceability and data structure
The final version of the SECP is applied to the actual Monaco Research Observatory corpus. Each spatial unit is identified by a Case ID; visual evidence by an Asset ID; and SECP decisions distinguish observation opportunities, confirmed occurrences, reviewed absences, and uncertain or non-assessable situations.
Spatial Cases
Visual Assets
Photographs
Videos
SECP-Candidate Assets
Confirmed Occurrences
Fieldwork Scope
The corpus brings together observations and visual evidence collected between 2021 y 2026 across heterogeneous spatial units in Monaco. The published results are based on the consolidated corpus and the traceable empirical decisions of the current version.
Evidence Architecture
Candidate-to-Occurrence Workflow
- Identify a potentially relevant feature, rule, interaction or transformation.
- Check whether there is a genuine opportunity to assess the corresponding SECP category.
- Distinguish contextual evidence from evidence sufficient to confirm a mechanism.
- Review alternative functional, legal, operational, security-related, privacy-related or maintenance-related explanations.
- Confirm the occurrence, record an assessed absence or preserve uncertainty, according to the observation opportunity and available evidence.
- Characterise confirmed occurrences using M1–M8 without multiplying them by the number of available assets.
Dataset Structure
The data architecture keeps the levels of case, asset, opportunity and occurrenceseparate. This prevents a candidate photograph from automatically becoming a confirmed mechanism and makes it possible to reconstruct the decision chain from the original evidence to the published result.
Unit of Assessment
Case ID × SECP Category
The unit used to construct comparative denominators is the combination of spatial case and SECP category. In this way, each pattern is assessed only where a genuine opportunity to observe it exists.
Observation Opportunities
900 case×category assessments
The 75 cases were cross-classified against the twelve SECP categories. Each combination was classified according to whether the mechanism could genuinely be assessed.
YES
A specific observable opportunity exists. It is included in the category denominator.
UNCERTAIN
A potential basis exists, but the evidence does not allow a sufficiently specific assessment.
NO
There is no relevant opportunity to assess this mechanism in the case. This status is not equivalent to a confirmed absence.
180 + 183 + 537 = 900. Within the 180 YES opportunities, the final matrix records 121 REAL_CONFIRMED decisions and 59 REAL_ABSENT decisions.
Normalization Rule
Confirmed occurrences ÷ observable opportunities
The normalised rate for each category is calculated only from Opportunity = YES. UNCERTAIN and NO combinations are excluded from the denominator.
Percentage = (confirmed occurrences ÷ YES opportunities) × 100. Percentages are rounded to one decimal place.
This normalisation enables a descriptive comparison conditional on observed opportunities. It does not make the samples for each category equivalent or remove differences in context or selection.
Occurrence-Level Coding
SE1–SE12 identify the mechanism · M1–M8 characterise the occurrence
Presence
Restrictive intensity
Explicitness
Spatial coverage
Temporal persistence
Behavioural effect
User specificity
Evidence confidence
Confirmation of a category does not follow automatically from the existence of a candidate photograph. The decision is made at mechanism level and must be supported by sufficient evidence for the corresponding case×category combination.
Internal Review & Reliability Scope
Internal review and category boundaries
The 180 assessments with Opportunity = YES were reviewed by the researcher responsible for the corpus, resulting in 121 REAL_CONFIRMED decisions and 59 REAL_ABSENT. These results are final for the published version. The review documents the application of operational rules and the traceability of decisions; it does not constitute an estimate of intercoder agreement.
Documented Internal Controls
- Application of the definitions and rules in the SECP Codebook.
- Specific review of positive, negative, uncertain and borderline cases.
- Retention of ambiguities and borderline decisions until their review.
- Resolution at mechanism and category level, rather than through an intuitive overall judgement.
- Separate recording of Evidence Confidence using M8.
- Rejection of secondary categories where no clearly distinct additional function exists.
Category-Boundary Clarifications
SE1 vs SE2 · Material feature vs route effect
SE1 is used when the analytical focus is the physical or architectural feature acting as a deterrent. SE2 is used when the main effect is to direct, divert or structure the circulation route.
SE4 vs SE11 · Symbolic signalling vs practical ambiguity
SE4 is reserved for visual or symbolic signals related to legitimacy, belonging or the social interpretation of space. SE11 is applied when the main issue is the practical ambiguity involved in recognising, locating or interpreting an accessible entrance or route.
SE8 vs SE9 · Surveillance vs staff mediation
SE8 requires observable surveillance or security presence together with contextual evidence of a restrictive function; the mere presence of a camera or staff member is insufficient to confirm soft exclusion. SE9 requires active human mediation: questioning, redirecting, delaying, conditioning or directly managing access or continued presence, with evidence of the corresponding filtering function.
Review Profile
The methodological review focuses on the choice of primary category, the distinction between sufficient and uncertain evidence, the intensity of restriction, assessment of M8 and the need for a secondary category. Final decisions retain traceability and apply the category-boundary rules.
The documented scope is an operational review of the corpus by a single coder. No formal intracoder reliability estimate through recoding or independent reproducibility assessment is presented. M8 expresses confidence in the evidence, not agreement between assessors.
Recommended procedures for studies involving multiple coders, and their distinction from the controls applied here, are explained in Reliability, Validity and Ethics.
Final Empirical Results
Confirmed mechanisms by observable opportunity
The percentages express the proportion of confirmed occurrences among cases in which the corresponding mechanism could be assessed empirically.
| SECP | Category | Opportunity N | Confirmed n | Normalized Rate | Visual |
|---|---|---|---|---|---|
| SE1 | Physical and Architectural Deterrence | 26 | 19 | 73.1% | |
| SE2 | Route Steering and Circulation Control | 28 | 17 | 60.7% | |
| SE3 | Furniture and Resting Restrictions | 12 | 8 | 66.7% | |
| SE4 | Visual and Symbolic Signalling | 25 | 17 | 68.0% | |
| SE5 | Signage and Informational Filtering | 16 | 12 | 75.0% | |
| SE6 | Economic Conditionality | 10 | 10 | 100.0% | |
| SE7 | Normative and Behavioural Filtering | 12 | 9 | 75.0% | |
| SE8 | Surveillance and Security Presence | 7 | 5 | 71.4% | |
| SE9 | Staff-Mediated Interactional Filtering | 16 | 9 | 56.3% | |
| SE10 | Temporal Exclusion | 12 | 7 | 58.3% | |
| SE11 | Spatial Ambiguity and Discreet Access | 9 | 5 | 55.6% | |
| SE12 | Temporary Privatisation and Event Transformation | 7 | 3 | 42.9% |
Interpretation: the rates represent confirmation within observable opportunities. They do not represent overall territorial prevalence or the percentage of affected visitors. Categories with small denominators should be interpreted with particular caution.
Comparative Reading
Patterns with the highest normalised rates
The comparison identifies which mechanisms were most frequently confirmed among the cases in which they could be assessed. This is not a ranking of social importance or territorial extent, nor does it demonstrate statistically significant differences between categories.
SE6 · Economic Conditionality
10 confirmations out of 10 observable opportunities.
SE5 · Signage and Informational Filtering
12 confirmations out of 16 observable opportunities.
SE7 · Normative and Behavioural Filtering
9 confirmations out of 12 observable opportunities.
SE1 · Physical and Architectural Deterrence
19 confirmations out of 26 observable opportunities.
SE8 · Surveillance and Security Presence
5 confirmations out of 7 observable opportunities.
SE4 · Visual and Symbolic Signalling
17 confirmations out of 25 observable opportunities.
Negative Cases & Alternative Explanations
Restriction is not automatically soft exclusion
Reviewing negative cases can rule out an interpretation of soft exclusion when an alternative explanation is better supported. Insufficient evidence, by contrast, may require uncertainty to be preserved: it does not by itself demonstrate absence.
Negative-Case Principle
The presence of a barrier, a high price, a luxury façade, a camera, staff, discreet signage or a temporary transformation does not in itself confirm an SECP category. The decision depends on the observable function of the mechanism, its context and the assessment of plausible alternative explanations.
| Candidate feature | Category assessed | Interpretive check |
|---|---|---|
| Bollards / physical barriers | SE1 | Check whether they actually restrict pedestrian movement or serve vehicle management. |
| Route diversion | SE2 | Distinguish selective route steering from universal diversions for safety or operational reasons. |
| Absence of seating | SE3 | Do not confirm without contextual evidence; the absence of furniture alone is insufficient. |
| Luxury aesthetics | SE4 | Luxury aesthetics do not prove exclusionary signalling or an effect on access. |
| Signage | SE5 | Assess clarity, placement, alternative information sources and practical effect. |
| High prices | SE6 | Distinguish the price of consumption from an actual restriction on circulation or use. |
| Behavioural rules | SE7 | Check the rule’s application, scope and functional justification. |
| Cameras / security | SE8 | The mere presence of surveillance does not confirm a restrictive effect. |
| Staff presence | SE9 | Require observable mediation with a filtering function, rather than mere presence or routine customer service. |
| Closure / schedule | SE10 | Distinguish temporal exclusion from ordinary published opening hours. |
| Discreet entrance | SE11 | Check whether genuine practical ambiguity exists at the decision point. |
| Event barrier | SE12 | Distinguish temporary privatisation from universal safety or emergency requirements. |
Uncertain Cases
The final structure retains UNCERTAIN as a separate status. The current corpus contains 183 case×category assessments with Opportunity = UNCERTAIN. These combinations are excluded from the normalised denominator and are not assigned absence, presence or a low score by default.
Situations that may justify uncertainty include insufficient visual evidence, non-comparable timing, an unobservable staff function, unverified legal status, a lack of contextual documentation, and the coexistence of equally plausible security, privacy or operational explanations.
Operational and Interpretive Controls
- Opportunity normalization: only Opportunity = YES is included in the denominator.
- Uncertainty preservation: UNCERTAIN is not forced into YES or NO.
- Asset–occurrence separation: multiple assets may provide additional evidence, but they do not multiply occurrences and are not presumed to be independent observations.
- Alternative-explanation review: safety, legal obligations, operations, privacy, maintenance, mobility and event logistics are reviewed before confirmation.
- Category-boundary control: explicit rules are applied for SE1/SE2, SE4/SE11 and SE8/SE9.
- VAM separation: a VAM profile does not automatically generate a SECP code.
Ethical & Interpretive Safeguards
- Social class or protected characteristics are not inferred from appearance.
- Behaviour is described in observable terms without attributing psychological states.
- The intentions of owners, designers or staff are not inferred from an isolated mechanism.
- Safety, privacy and operational explanations remain genuine alternatives.
- Evidence is linked through traceable identifiers, not through the reputation of a location.
Review Principle
The review confirms mechanisms when sufficient evidence exists, preserves uncertainty when it cannot be resolved, and records assessed absences when the evidence supports that decision. These situations must remain distinct.
Empirical Interpretation
Diversity of confirmed mechanisms within the corpus
The results document material, informational, economic, normative, visual and surveillance-related mechanisms. Among observable opportunities, the highest rates are found for economic conditionality, signage and information, normative and behavioural filtering, physical deterrence, surveillance and visual or symbolic signalling.
This diversity is consistent with interpreting soft exclusion as a spatial and situational phenomenon. However, the table aggregated by category does not by itself demonstrate that mechanisms coexist or interact within individual cases. Such an interpretation requires examination of their joint distribution in the case matrix.
Lower rates in other categories do not imply lower analytical relevance. Mechanisms have different denominators and opportunity structures and, in some cases, small sample sizes.
VAM–SECP Relationship
Shared evidence · Separate analytical decisions
The Visibility–Accessibility Matrix and SECP may draw on the same visual evidence, but they address different questions. VAM describes visibility and accessibility configurations through V1–V5 and A1–A5; SECP identifies mechanisms through SE1–SE12 and characterises them using M1–M8.
SECP does not generate VAM scores
Shared evidence · Separate coding decisions
This separation is analytical: it does not imply statistical independence between the two instruments, nor does it constitute, by itself, evidence of convergent or discriminant validity.
Interpretative Limits
What these results allow us to conclude
They allow confirmed SECP mechanisms to be described and compared within the observable opportunities of the corpus. The case matrix allows their distribution across units and types of space to be examined.
They are not equivalent to the percentage of territory affected, the percentage of visitors excluded or a direct measure of subjective experience. The rates describe the corpus and must not be automatically generalised to Monaco as a whole.
Simply adding the categories present does not automatically constitute a scale of exclusion intensity. Nor do differences between rates demonstrate causality or statistical significance.
The final status of the corpus is compatible with these limitations. Review by a single coder does not allow independent agreement to be estimated. This page also does not report separate studies of content validity involving an external panel, convergent validity, discriminant validity or temporal validity.
Validation Scope & Conclusion
Final corpus results · Documented operational application
The empirical application documents the use of the twelve categories as an operational system for identifying and comparing mechanisms within the corpus studied. Final decisions are linked to observable opportunities and traceable evidence, with an internal review conducted by the researcher responsible.
The main result is a comparative profile of confirmed mechanisms conditioned by their observability structure, not a general exclusion rate for Monaco.
This conclusion is limited to the documented application and review. It does not establish intercoder agreement or replace independent reliability assessments or specific validity studies.
Research Connections
Continue the analysis
Soft Exclusion Coding Protocol
SECP Overview · Coding Taxonomy · Observation & Coding Method · Reliability, Validity and Ethics · Research Materials · Empirical Validation & Final Results
