Plan and analyse a collection survey

This tool helps you with the statistics to plan, carry out and interpret collection surveys. It accompanies the booklet "Introduction to statistics for sample surveys in heritage collections" by Cristina Duran Casablancas, Hilde Schalkx, Scott Allan Orr and Josep Grau-Bové. The app is now in Beta version, feedback is welcome.

Sample Size Calculator

Calculate the target sample size you need to estimate a mean or a proportion within a specific margin of error.

Launch Calculator

Confidence Interval Calculator

Calculate confidence intervals and margins of error from your survey results.

Launch Calculator

Sampling List Generator

Generate a random list of object IDs for sampling from a collection. Useful at the start of a survey.

Launch Calculator

Summary Stats from Data

Upload or paste spreadsheet columns to quickly obtain statistics and standard errors, including for logarithmic pH measurements.

Launch Calculator

Strata Sample Size Planning

Plan stratified sampling for different sections of a collection, using proportional or Neyman allocation.

Launch Calculator

Strata Confidence Intervals

Analyse stratified survey data and calculate confidence intervals.

Launch Calculator

1 Sample Size & Margin of Error

I want to estimate:

A mean of a population A proportion of a population

I have:

A fixed sample size A desired margin of error

My population size ($N$) is:

Unknown, but very large (at least $20n$) Known (specify below)

Population Size ($N$)

Expected Standard Deviation ($SD$) inside population

Expressed in the same units as the estimated mean.

Target Sample Size ($n$)

Confidence Level (typically 0.95)

Finite Population Correction (FPC) Engine Apply correct FPC factor to Margin of Error output when $N$ is known (Recommended)

Calculated Outcome

0.00

Distribution: z-distribution (default)

Understanding Standard Error & $MoE$

The Standard Error ($SE$) measures how much your estimate would vary if the survey were repeated many times. It indicates the statistical precision of your calculated mean or proportion.

The Margin of Error ($MoE$) defines the half-width of the confidence interval around your estimate ($\text{Estimate} \pm MoE$). It is calculated as:

$$MoE = \text{Critical Value } (z \text{ or } t) \times SE$$

2 CI Configuration

Estimate the Confidence Interval (CI) of:

A mean of a population A proportion of a population

My population size ($N$) is:

Unknown/Extremely Large Known (specify below)

Observed Sample Size ($n$)

Sample Estimated Mean

Sample Standard Deviation ($SD$)

Expressed in units of the mean.

Target Confidence Level (typically 0.95)

Results

Metric	Value

*Note: When sample size is lower than 30 ($n < 30$), the system defaults to a Student's $t$-statistic boundary rather than a $z$-statistic, matching standard sampling methodologies.

3 Random Item Selection Generator

Starting Number (Minimum ID)

Finishing Number (Maximum ID)

Required Sample Size (Amount to extract)

Generated Sampling Sequence

Click 'Shuffle & Generate' to compile your random survey IDs list.

4 Import Datasets

Option A: Upload Data File

Upload CSV, TSV or Plain text data

First column read, other columns disregarded.

Option B: Paste Data Directly

File Has Headers Logarithmic pH Mean

Population Size

Unknown Known:

Confidence Tier

Import Preview (First 6 Rows)

No active dataset parsed yet. Use demo button or upload one.

Analysis Findings

Evaluated Parameter	Value

Dataset Histogram

5 Stratified Sample Planning

Estimate:

A mean value A proportion value

Given Metric Type:

Fixed sample size ($n$) Target margin of error

Total Target Sample Size ($n$)

Stratification Allocation Rule:

Proportionate (Simple):

Select this if standard deviation of each stratum is identical or unknown.

Disproportionate (Neyman Allocation):

Select this if standard deviations across strata differ significantly to optimize statistical weight.

Configure Stratum Cohorts

Stratum Name	Pop. Size $N_h$	Expected SD	Action

Optimal Allocation Output

Segment Category	Target Sample Allocation

*Allocations are automatically rounded up (ceiling rules) to safeguard target limits.

6 Analyse Stratified Sampling data

Estimate Stratum CI of:

A mean value A proportion value

Confidence Level (CL)

Observed Strata Metric Parameters

Stratum Name	Pop. Size $N_h$	Sample Size $n_h$	Mean / Prop	SD	Action

Synthesized Population Results

Statistical Coefficient	Computed value