When a program release needs to be backed out of production, the changes introduced by that release must be removed from the source code to ensure the system returns to its prior state. This approach ensures that the source code reflects the stable version without the problematic changes.
References
* ISACA CISA Review Manual 27th Edition, Page 244-245 (Change Management)

Hashing is a technique that transforms data into a fixed-length value, called a hash or a digest, that uniquely represents the original data. Hashing can be used to validate the integrity of data communicated between production databases and a big data analytics system by comparing the hash values of the data before and after the communication. If the hash values match, the data has not been altered; if they differ, the data has been tampered with or corrupted. Hashing is a better security control than encrypting, running and comparing the count function, or hosting a digital certificate for this purpose because:
Encrypting in-scope data sets can protect the confidentiality of the data, but not necessarily the integrity.
Encryption algorithms can be broken or bypassed by malicious actors, or encryption keys can be compromised or lost. Moreover, encryption adds overhead to the communication process and may affect the performance of the big data analytics system.
Running and comparing the count function within the in-scope data sets can only verify the number of records or elements in the data sets, but not the content or quality of the data. The count function cannot detect any changes or errors in the data values, such as missing, duplicated, corrupted, or manipulated data.
Hosting a digital certificate for in-scope data sets can provide authentication and non-repudiation for the data sources, but not integrity for the data itself. A digital certificate is a document that contains information about the identity and public key of an entity, such as a person, organization, or device. A digital certificate does not contain or verify the actual data that is communicated between production databases and a big data analytics system.
References:
Ensuring Data Integrity with Hash Codes
Database Security: An Essential Guide
Control methods of Database Security

A digital signature is the most appropriate control to ensure integrity of online orders because it provides a way to verify the authenticity and integrity of the data sent by the sender. A digital signature is created by applying a cryptographic algorithm to the data and attaching the result to the data. The receiver can then use the sender's public key to verify that the data has not been altered or tampered with during transmission. A digital signature also provides non-repudiation, which means that the sender cannot deny sending the data.
Data Encryption Standard (DES) is a symmetric encryption algorithm that can provide confidentiality of online orders, but not integrity. DES uses the same key to encrypt and decrypt the data, which means that anyone who has the key can modify the data without detection.
Public key encryption is an asymmetric encryption algorithm that can also provide confidentiality of online orders, but not integrity. Public key encryption uses a pair of keys: a public key and a private key. The sender encrypts the data with the receiver's public key, and the receiver decrypts it with their own private key.
However, public key encryption does not prevent anyone from modifying the encrypted data.
Multi-factor authentication is a control that can provide authentication and authorization of online orders, but not integrity. Multi-factor authentication requires the user to provide two or more pieces of evidence to prove their identity, such as a password, a token, or a biometric factor. Multi-factor authentication can prevent unauthorized access to online orders, but it does not protect the data from being modified after being sent.
References:
* ISACA, CISA Review Manual, 27th Edition, 2019, p. 281 1
* ISACA, CISA Review Questions, Answers & Explanations Database - 12 Month Subscription 2

The most conclusive audit procedure for evaluating the effectiveness of an e-commerce application system's edit routine is to use test transactions. A test transaction is a simulated input that is processed by the system to verify its output and performance1. By using test transactions, an auditor can directly observe how the edit routine checks the validity, accuracy, and completeness of data entered by users, and how it handles incorrect or invalid data. A test transaction can also help measure the efficiency, reliability, and security of the edit routine, as well as identify any errors or weaknesses in the system.
The other options are not as conclusive as using test transactions, as they rely on indirect or secondary sources of information. Reviewing program documentation is an audit procedure that involves examining the written description of the system's design, specifications, and functionality2. However, program documentation may not reflect the actual implementation or operation of the system, and it may not reveal any discrepancies or defects in the edit routine. Interviews with knowledgeable users is an audit procedure that involves asking questions to the people who use or manage the system3. However, interviews with knowledgeable users may not provide sufficient or objective evidence of the edit routine's effectiveness, and they may be influenced by personal opinions or biases. Reviewing source code is an audit procedure that involves analyzing the programming language and logic of the system4. However, reviewing source code may not be feasible or practical for complex or large systems, and it may not demonstrate how the edit routine performs in real scenarios.

The best way to ensure that potential security issues are considered by the development team as part of incremental changes to agile-developed software is to include a mandatory step to analyze the security impact when making changes. This will help to identify and mitigate any security risks or vulnerabilities that may arise from the changes, and to ensure that the software meets the security requirements and standards. The other options are not as effective, because they either delegate the security analysis to someone outside the development team, rely on post-deployment testing, or focus on documentation rather than analysis. References: CISA Review Manual (Digital Version)1, Chapter 4, Section 4.2.5