Equivalent Text: A Deep Look into Sample Representativeness (Random Selection)

This static document mirrors the app’s behavior (same population and sampling logic) and presents accessible summaries of samples and cumulative results.

Introduction

Purpose. This document is an accessible, text-first companion to A Deep Look into Sample Representativeness When Using Random Selection. It mirrors the app’s logic without graphics so learners using screen readers or preferring text can explore random selection, representativeness, and how sample size affects results.

What you can learn. By examining ten independent samples at several sizes (n = 5, 25, 100, 1000), you can observe how sample-to-sample variation changes and how close sample means tend to be to the population mean (μ = 24). You’ll also see how the sampling distribution of sample means stabilizes as more samples accumulate.

What you will see for each size.

  1. A Sampling of Ten Samples — For each of the ten samples, we report Mean, Stdev, Median, IQR, Skewness, Min, Max, and the absolute deviation of the sample mean from μ (|x̄ − μ|).
  2. Representativeness & CLT (Cumulative Sample Means) — As we accumulate more samples, we summarize Number of Samples, Mean of Means, Standard Error (SD of sample means), Skewness of Means, and the best / worst observed |x̄ − μ| so far.

How to navigate. Use the Table of Contents or jump by headings. Each sample size appears under its own H2 heading.

All values are formatted to three decimals; integer counts use thousands separators.

Size 5

A Sampling of Ten Samples
MeanStdevMedianIQRSkewnessMinMax|x̄ − μ|Size
22.6006.50422.0009.0000.10315311.4005
24.8005.07026.0005.0000.19519320.8005
23.4005.89923.0008.000-0.24915290.6005
21.2004.55023.0006.000-0.28215262.8005
22.4005.03022.0009.0000.04817281.6005
25.2005.35727.0005.000-0.10418321.2005
25.4006.73127.0007.000-0.49215321.4005
23.0006.81925.0001.000-0.98411281.0005
23.6004.27823.0008.0000.06719280.4005
22.8007.66225.0006.000-0.76010291.2005

Each row summarizes one independently drawn sample of size 5 from the SWLS population (scores 5–35). The |x̄ − μ| column shows how far the sample mean is from the population mean (μ = 24). Values are reported to three decimals.

Representativeness & CLT (Cumulative Sample Means)
Number of SamplesMean of MeansStandard ErrorSkewness of MeansBest |x̄ − μ|Worst |x̄ − μ|Sample Size
129.2000.000NaN5.2005.2005
227.6002.263NaN2.0005.2005
324.9334.888-0.9352.0005.2005
425.0503.998-0.9341.4005.2005
525.3203.515-1.2211.4005.2005
625.4673.164-1.4071.4005.2005
725.9433.151-1.4721.4005.2005
825.2503.514-0.8131.4005.2005
924.5783.857-0.4591.4005.2005
1024.6003.637-0.4960.8005.2005
2024.6203.175-0.7350.2006.0005
3024.2333.179-0.5200.0006.8005
4024.2853.011-0.5100.0006.8005
5024.0802.918-0.4240.0006.8005
6024.2232.788-0.5040.0006.8005
7024.1402.692-0.4190.0006.8005
8023.9832.710-0.3460.0006.8005
9023.9222.926-0.2900.0006.8005
10023.8162.907-0.2390.0006.8005
20023.7242.845-0.2440.0009.6005
30023.7092.731-0.2180.0009.6005
40023.7712.681-0.1920.0009.6005
50023.8742.625-0.2260.0009.6005
60023.8862.645-0.2200.0009.6005
70023.9192.689-0.2250.0009.6005
80023.9092.718-0.2350.0009.6005
90023.9042.707-0.1920.0009.6005
1,00023.8922.709-0.1940.0009.6005

Cumulative results after each of the first 10 samples, then every 10 through 100, and every 100 through 1,000. “Best |x̄ − μ|” is the smallest absolute deviation from the population mean (μ = 24) observed so far; “Worst |x̄ − μ|” is the largest.

Size 25

A Sampling of Ten Samples
MeanStdevMedianIQRSkewnessMinMax|x̄ − μ|Size
23.3206.94524.0006.000-0.6367340.68025
23.6006.57025.00010.000-0.4827330.40025
23.4005.40123.0007.0000.11315330.60025
23.2006.51925.00010.000-0.10211350.80025
22.4006.56423.0006.000-0.5938341.60025
23.6405.86623.0006.000-0.06511330.36025
22.9205.72923.0006.000-0.4319331.08025
23.7206.18124.0008.0000.14613340.28025
25.0805.75825.0007.000-0.45613351.08025
24.4405.73125.0006.000-0.62610350.44025

Each row summarizes one independently drawn sample of size 25 from the SWLS population (scores 5–35). The |x̄ − μ| column shows how far the sample mean is from the population mean (μ = 24). Values are reported to three decimals.

Representativeness & CLT (Cumulative Sample Means)
Number of SamplesMean of MeansStandard ErrorSkewness of MeansBest |x̄ − μ|Worst |x̄ − μ|Sample Size
123.4800.000NaN0.5200.52025
224.2601.103NaN0.5201.04025
324.2930.782-0.3810.3601.04025
423.9001.013-0.0890.3601.28025
524.1681.063-0.5070.3601.28025
623.8131.288-0.2820.3601.96025
723.8511.180-0.4180.0801.96025
824.1051.307-0.3260.0801.96025
924.0001.262-0.0510.0801.96025
1023.9441.2030.1100.0801.96025
2023.8301.3580.4710.0802.44025
3023.9961.2180.2810.0402.44025
4024.0451.1830.2700.0002.44025
5024.1421.2630.2880.0003.24025
6024.0311.2140.4050.0003.24025
7024.0561.2120.3840.0003.24025
8024.0911.1900.3250.0003.24025
9024.0401.1570.3920.0003.24025
10024.0411.1580.4270.0003.24025
20024.1231.128-0.0450.0003.36025
30024.0931.134-0.1030.0003.36025
40024.0251.139-0.1180.0003.52025
50023.9691.147-0.1150.0004.00025
60023.9561.158-0.1030.0004.00025
70023.9651.162-0.0640.0004.00025
80023.9751.181-0.0530.0004.00025
90023.9791.176-0.0930.0004.00025
1,00023.9941.185-0.0840.0004.00025

Cumulative results after each of the first 10 samples, then every 10 through 100, and every 100 through 1,000. “Best |x̄ − μ|” is the smallest absolute deviation from the population mean (μ = 24) observed so far; “Worst |x̄ − μ|” is the largest.

Size 100

A Sampling of Ten Samples
MeanStdevMedianIQRSkewnessMinMax|x̄ − μ|Size
23.8506.05124.0008.500-0.20510340.150100
23.9705.70224.0008.250-0.03110350.030100
24.4905.58024.0009.0000.05612350.490100
23.3005.76624.0006.250-0.4517350.700100
24.1005.51924.0008.000-0.10411340.100100
23.6905.84424.0009.000-0.23010350.310100
23.2106.45623.0008.000-0.4836340.790100
23.4206.61123.00011.250-0.1506340.580100
24.5005.34425.0007.000-0.39112340.500100
24.6506.26725.0008.250-0.2309350.650100

Each row summarizes one independently drawn sample of size 100 from the SWLS population (scores 5–35). The |x̄ − μ| column shows how far the sample mean is from the population mean (μ = 24). Values are reported to three decimals.

Representativeness & CLT (Cumulative Sample Means)
Number of SamplesMean of MeansStandard ErrorSkewness of MeansBest |x̄ − μ|Worst |x̄ − μ|Sample Size
124.7100.000NaN0.7100.710100
224.1100.849NaN0.4900.710100
323.8470.7541.6090.4900.710100
424.0070.6940.0220.4900.710100
523.7320.861-0.0310.4901.370100
623.7050.7730.1230.4301.370100
723.7210.7070.0290.1801.370100
823.8400.735-0.2220.1801.370100
923.9330.743-0.4370.1801.370100
1023.9330.700-0.4490.0701.370100
2023.9960.607-0.2400.0601.370100
3024.1410.6570.2830.0401.860100
4024.1400.6080.2240.0201.860100
5024.1050.6060.0080.0201.860100
6024.0890.6100.0960.0201.860100
7024.0750.5870.1130.0201.860100
8024.0400.5840.1220.0201.860100
9024.0310.5760.0960.0201.860100
10024.0460.5710.0810.0201.860100
20024.0340.5760.1610.0101.860100
30024.0320.5910.1210.0101.860100
40024.0270.607-0.0210.0101.880100
50024.0100.601-0.0370.0001.880100
60024.0220.591-0.0390.0001.880100
70024.0230.591-0.0000.0001.880100
80024.0290.606-0.0390.0001.880100
90024.0240.615-0.0180.0001.880100
1,00024.0230.609-0.0190.0001.880100

Cumulative results after each of the first 10 samples, then every 10 through 100, and every 100 through 1,000. “Best |x̄ − μ|” is the smallest absolute deviation from the population mean (μ = 24) observed so far; “Worst |x̄ − μ|” is the largest.

Size 1000

A Sampling of Ten Samples
MeanStdevMedianIQRSkewnessMinMax|x̄ − μ|Size
23.9376.09424.0009.000-0.3715350.0631,000
23.9895.82424.0008.000-0.3115350.0111,000
24.1935.84024.0009.000-0.3536350.1931,000
24.0586.10224.0009.000-0.3886350.0581,000
23.9075.94024.0008.000-0.4076350.0931,000
24.0825.97425.0009.000-0.4366350.0821,000
23.9596.23725.0009.000-0.4265350.0411,000
24.1736.06824.0009.000-0.3146350.1731,000
23.8836.12124.0009.000-0.3125350.1171,000
24.2776.00525.0008.000-0.4065350.2771,000

Each row summarizes one independently drawn sample of size 1,000 from the SWLS population (scores 5–35). The |x̄ − μ| column shows how far the sample mean is from the population mean (μ = 24). Values are reported to three decimals.

Representativeness & CLT (Cumulative Sample Means)
Number of SamplesMean of MeansStandard ErrorSkewness of MeansBest |x̄ − μ|Worst |x̄ − μ|Sample Size
123.9060.000NaN0.0940.0941,000
223.8520.076NaN0.0940.2021,000
323.8570.055-0.7950.0940.2021,000
423.9300.1541.4940.0940.2021,000
523.9280.1331.5270.0820.2021,000
623.8840.1600.5900.0820.3331,000
723.9030.1540.1400.0170.3331,000
823.9130.145-0.1020.0170.3331,000
923.9270.142-0.3610.0170.3331,000
1023.9250.134-0.3040.0170.3331,000
2023.9540.1500.0880.0170.3331,000
3023.9800.164-0.1410.0170.3331,000
4023.9970.1660.0460.0130.3821,000
5023.9970.1730.0630.0130.3821,000
6024.0020.175-0.0260.0130.3821,000
7023.9910.1730.0520.0130.3821,000
8024.0010.1730.0610.0130.3821,000
9024.0070.1680.0010.0110.3821,000
10024.0000.1700.0390.0110.3821,000
20024.0110.178-0.0640.0020.4551,000
30024.0070.1780.0480.0010.4761,000
40024.0010.1810.1230.0000.5861,000
50024.0040.1830.0030.0000.5861,000
60023.9980.186-0.0060.0000.5861,000
70023.9950.1890.0510.0000.6441,000
80024.0010.1890.0220.0000.6441,000
90024.0030.1870.0260.0000.6441,000
1,00024.0010.1860.0480.0000.6441,000

Cumulative results after each of the first 10 samples, then every 10 through 100, and every 100 through 1,000. “Best |x̄ − μ|” is the smallest absolute deviation from the population mean (μ = 24) observed so far; “Worst |x̄ − μ|” is the largest.

Methodology for Generating Values

Population. SWLS scores are integers from 5 to 35. We simulate a truncated normal so that the truncated population has mean μ = 24 and SD = 6. We solve for parameters (μsim, σsim) of an underlying normal and truncate to [5,35].

Sampling. Each sample of size n is drawn independently with the same truncation and integer rounding used in the app.

Statistics. Per-sample: Stdev is the sample SD (n−1); skewness is m3/n divided by SD³ (uncorrected). For the cumulative table (sample means), skewness is the bias‑corrected estimator. Median and IQR use linear interpolation (Hyndman–Fan Type 7) when needed. Standard Error is the sample SD of sample means.