Performing statistical calculations is essential for quantity surveyors in Kenya when analyzing project data such as cost estimates, material quantities, and labour productivity. Accurate identification, collection, and organization of data ensure reliable statistical analysis supporting decisions on budgeting, tender evaluation, and cost control. This chapter focuses on mastering these preliminary steps to handle data effectively before applying statistical tools.

2.1 Identification, Collection and Organization of Data

In quantity surveying, data must be carefully identified and collected from project sites, contracts, and procurement records to ensure consistency and accuracy. Organizing data systematically allows for efficient retrieval and analysis, which supports cost estimation and risk assessment. This section covers the mathematical approach to handling data for statistical analysis.

2.1.1 Identification of Data

Data identification involves recognizing relevant quantitative information from project documents, site measurements, and financial records. In quantity surveying, this includes measurements of materials, labour hours, unit costs, and quantities of work completed. The type of data can be classified as discrete or continuous depending on the nature of the measurement.

The key mathematical concept is understanding data types for proper statistical treatment:

• Discrete data: Countable values such as number of bricks, quantity of cement bags.
• Continuous data: Measurable quantities like length in meters, weight in kilograms.
• Nominal data: Categories without order, e.g., types of materials.
• Ordinal data: Categories with order, e.g., quality ratings from low to high.

Worked Examples

Example 1: Identify data type for number of bricks used in a wall construction.

Given: Quantity of bricks = 1500 bricks

Discrete data because bricks are countable individual units.

Example 2: Identify data type for thickness of plaster applied on walls measured at 12 mm.

Given: Thickness = 12 mm

Continuous data because thickness is a measurable quantity with decimals.

Example 3: Identify data type for classifying cement quality as Grade 42.5.

Given: Cement grade = 42.5

Ordinal data as grades have ranking based on strength.

2.1.2 Collection of Data

Data collection in quantity surveying involves systematic gathering of measurements and records, ensuring accuracy and consistency. The process typically follows these steps:

1. Define purpose of data collection (e.g., cost estimation, progress tracking).
2. Identify sources (site records, invoices, supplier catalogs).
3. Select data collection tools (measurement tapes, digital devices, software).
4. Record data using standard formats (measurement sheets, spreadsheets).
5. Verify data accuracy through cross-checking.
6. Store data securely for analysis.

Worked Examples

Example 1: Calculate total material quantity collected from site records.

Given: 250 bags of cement, 1200 bricks, 50 steel bars

Calculate total items collected:

Total = 250 + 1200 + 50 = 1500 items

Example 2: Collect daily labour hours for a week: 8, 9, 7.5, 8, 8.5, 7, 8 hours.

Calculate total labour hours:

Total hours = 8 + 9 + 7.5 + 8 + 8.5 + 7 + 8 = 56 hours

Example 3: Collect measurements of wall length: 4.5 m, 5.2 m, 3.8 m, 4.9 m.

Calculate average wall length:

Average length = (4.5 + 5.2 + 3.8 + 4.9) / 4 = 18.4 / 4 = 4.6 m

2.1.3 Organization of Data

Organizing data involves arranging it logically for easy analysis, often using frequency distributions, tables, or spreadsheets. In quantity surveying, data is organized by project phase, material type, or cost category. Proper organization facilitates calculating measures like mean, median, and mode.

The mathematical formula for frequency \( f \) is:

$$ f = \frac{\text{Number of occurrences of a value}}{\text{Total number of data points}} $$

Worked Examples

Example 1: Organize the following quantities into a frequency table:

Quantities: 10, 12, 10, 15, 12, 10, 15, 15, 12, 10

Calculate frequency for quantity 10:

$$ f = \frac{4}{10} = 0.4 $$

Example 2: Organize labour hours into intervals and find frequency:

Labour hours: 6, 7, 8, 7, 9, 6, 8, 7, 9, 6

Intervals: 6-6.9, 7-7.9, 8-8.9, 9-9.9

Example 3: Organize cost data in Ksh for material purchases:

Costs: 1200, 1500, 1200, 1700, 1500, 1200, 1700

Calculate frequency for cost 1500:

$$ f = \frac{2}{7} \approx 0.286 $$

2.1.4 Data Validation and Cleaning

Data validation ensures correctness and completeness before statistical calculations. In quantity surveying, errors like missing entries or inconsistent units can distort results. Cleaning involves detecting and correcting such errors.

Steps for data validation and cleaning:

1. Check for missing values in datasets.
2. Verify consistency of units (e.g., meters, kilograms).
3. Identify outliers that deviate significantly from others.
4. Correct or remove erroneous entries.
5. Confirm data alignment with source documents.
6. Prepare clean dataset for analysis.

Worked Examples

Example 1: Validate quantity data with missing entries:

Data: 100, 120, -, 110, 115

Step 1: Identify missing value (-).

Step 2: Estimate missing value as average of known values.

Average = (100 + 120 + 110 + 115) / 4 = 111.25

Replace missing entry with 111.25

Example 2: Detect unit inconsistency in length data:

Data: 3.5 m, 350 cm, 4.2 m

Convert 350 cm to meters:

$$ 350 \text{ cm} = \frac{350}{100} = 3.5 \text{ m} $$

All data now consistent in meters.

Example 3: Identify outlier in labour hours:

Data: 8, 7.5, 8, 20, 7.8

Calculate mean:

$$ \text{Mean} = \frac{8 + 7.5 + 8 + 20 + 7.8}{5} = \frac{51.3}{5} = 10.26 $$

20 is significantly higher than others; investigate and correct if erroneous.

Chapter Summary

This chapter began with methods for identifying, collecting, and organizing data effectively to ensure accuracy and relevance in analysis. It then covered the interpretation, analysis, and presentation of data using appropriate formats to communicate findings clearly. Measures of central tendency including mean, median, and mode, along with standard deviation as a measure of dispersion, were thoroughly explored through calculations. The chapter progressed to laws of probability and their application in calculating probability distributions, mathematical expectations, and sampling distributions. Various sampling distribution methods were examined for their role in data analysis, followed by calculations using the standard normal table, T-distribution, and estimation techniques. Confidence intervals were introduced to quantify the reliability of estimates, and hypothesis testing was explained using procedures for both large and small samples. Finally, the chapter concluded with calculations involving correlation, regression, rank correlation coefficients, and equations of the regression line, integrating these statistical tools for comprehensive data analysis.